The International Simutrans Forum

 

Author Topic: PAK128 Pricing & Balancing - Discussion  (Read 58953 times)

0 Members and 1 Guest are viewing this topic.

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #70 on: October 22, 2008, 08:12:31 AM »
zeno: If you need some help / a second person for testing the new cost structure, let me know - I can test them (now and later versions as well).

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #71 on: October 22, 2008, 08:33:13 AM »
zeno: If you need some help / a second person for testing the new cost structure, let me know - I can test them (now and later versions as well).
Nice! I'll let you know when I finish with wagons. Then some testing on train routes will be needed.

Offline wernieman

  • Devotees (Inactive)
  • *
  • Posts: 713
    • Werniemans-Webside (only German)
Re: PAK128 Pricing & Balancing - Discussion
« Reply #72 on: October 22, 2008, 11:51:55 AM »
If you have a "beta", we could make a "spezial-nightly-pak128"
I hope you understand my English

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #73 on: October 22, 2008, 12:09:46 PM »
If you have a "beta", we could make a "spezial-nightly-pak128"
Thanks werner! Later on it can be a good idea, I'll tell you when ready. By now, too much testing needed until an acceptable version is ready :P

Offline wernieman

  • Devotees (Inactive)
  • *
  • Posts: 713
    • Werniemans-Webside (only German)
Re: PAK128 Pricing & Balancing - Discussion
« Reply #74 on: October 22, 2008, 01:36:59 PM »
P.S. if you need, I could make a spezial SVN .....

This is only a information!
I hope you understand my English

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #75 on: October 22, 2008, 02:46:29 PM »
P.S. if you need, I could make a spezial SVN .....
Thanks, werner, but IMHO no SVN is needed: the only new thing I manage is the excel sheet, so a special SVN for it I think is too much ;)

By the way, I've got the first "beta" of the pak (actually I only created the vehicles.all.pak), but I don't know wether a special nightly for pak128 is needed. It should be a decision of the PAK admins. It's good to know you can provide the service; and I can provide the product; so now we know it can be created... It's just a decision matter.

Well, I'll begin to test it as soon as I arrive home!


Offline VS

  • Senior Plumber (Devotee)
  • Moderator
  • *
  • Posts: 4852
  • Vladimír Slávik
    • VS's Simutrans site
  • Languages: CS,EN
Re: PAK128 Pricing & Balancing - Discussion
« Reply #76 on: October 22, 2008, 03:06:57 PM »
From my side there is absolutely no objection to releasing testing versions.

Maybe making the first results available could provide feedback from the players?

My projects... Tools for messing with Simutrans graphics. Graphic archive - templates and some other stuff for painters. Development logs for most recent information on what is going on. And of course pak128!

Offline wernieman

  • Devotees (Inactive)
  • *
  • Posts: 713
    • Werniemans-Webside (only German)
Re: PAK128 Pricing & Balancing - Discussion
« Reply #77 on: October 22, 2008, 03:19:01 PM »
if you have make a vehicles.all.pak ... then you cold make it available for player like an alpha-Pak

When a player have:
nightly-simutrans
nightly-PAK128
Your "aplha" PAK
.... The he could give you better feedback  :police:
I hope you understand my English

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #78 on: October 22, 2008, 03:28:25 PM »
Yes, it's a great idea to adquire feedback via users using a nightly.
If you want, I can provide the vehicles.all.pak, or the modified DAT files, or I can try to compile the whole pak...
Tell me the best way and what files should I update... and we'll have a new nightly :D

Offline wernieman

  • Devotees (Inactive)
  • *
  • Posts: 713
    • Werniemans-Webside (only German)
Re: PAK128 Pricing & Balancing - Discussion
« Reply #79 on: October 22, 2008, 03:34:27 PM »
When you only change the vehicle then make only the vehicle.all.pak

When you make bigger changes, then I could make a spezial-nightly-pak ;o)

For an alpha pak it is better to make only parts of the pak, so only better informed user used them.
(I hope nobody misunderstand me)
I hope you understand my English

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #80 on: October 22, 2008, 09:30:29 PM »
Sorry, I messed up with some things in the last post... I've just deleted and rewritten it :-[

Now I've done first (and second) tests... And here are the results:
I used a saved game; date was 22nd november 1971 and I used each different compiled pak until reach end of year. First column is end of year train's benefit with the old pak, and the following are the same with the altered paks (modifications are indicated in column header; 10%income - 55%bonus, 10%income - 100%bonus, and 15%income - 100%bonus).

Conclusions:
I don't see a really important changes despite changing the parameters. Really the results are what I expected, but in a very small scale: I thought the impact would be much bigger. 
I can see a big change is noticed in car transport. Notice that all tracks for car transport are 110kmh, despite the engine, and they are all using long car wagon (by Rvg), which have 6 cars capacity and its RC is 0.13cr/km. I'll have to take a look at this (deeply).

Edit: I forgot to say (almost) all goods trains were 8 squares long, 14/15 wagons for cargo.
« Last Edit: October 22, 2008, 09:39:35 PM by Zeno »

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #81 on: October 23, 2008, 12:19:22 PM »
I am sorry I do not understand what is the default in the above tables = what does the first column (acum. year benefit ...) correspond to (to which setting)?

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #82 on: October 23, 2008, 01:40:16 PM »
NOTE: This concerned version with 10-year-payback prices as opposed to 5-year-payback prices. Clarification will follow.
It seems though that even with the latest (5-year) prices the text below applies.
--------------------------------------------------------------------------------------------------------------------

By the way, I have shortly experimented with the new prices for trains transporting goods (not passangers) in 1930s (one of the typical starting period for a game). In various realistic scenarios for raw materials such as coal, which belong to initially transported goods at the start of a game, and for various steam locs it seems that the vehicles are really paid back by their profits roughly within 10 year provided that (i) a reasonable length of convoys is chosen (not too short = low capacity, not too long = long speed), (ii) the train is on average 50% full, and (iii) train does not stand/wait anywhere too long. On the other hand, it is not too difficult to get some profit with most trains as long as they are operating close to their 50% average capacity and there not many unnecessary maintenance items.

This means that the 10-years return on investment is achieved only at the start of a game only if one is playing wisely, which is OK. On the other hand, setting up a train transport from one factory to another can consume easily 100000-150000 given the higher purchasing costs (depending on the distance etc.). In my opinion, we could thus possibly increase the starting capital to 500000 so that about three train connection (separate or a chain) can be built up in initial years - otherwise, the gameplay could become too boring in the first years. (One can of course argue that cars are preferable at the beginning, but their purchasing prices will probably adjust as well...).


P.S.: I like the new prices also because the ratio purchase / running cost gets a bit more realistic even though they might make the play slightly more difficult at the beginning - but that is something to be decided, that is, how difficult should pak128 be?

« Last Edit: October 23, 2008, 02:16:24 PM by jatypc »

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #83 on: October 23, 2008, 01:47:21 PM »
I am sorry I do not understand what is the default in the above tables = what does the first column (acum. year benefit ...) correspond to (to which setting)?
Well, it has not many sense there; It was originally put there to have a reference, but now it may be nonsense. That column has the values of the vehicle's benefit on 22nd september 1971. It was before changing anything, standard pak128. The following columns are all the benefit at 31st december 1971 with several configurations.
Sorry, I hope you understand now  :P



I'm agree with all of your last post, but take care with the payback period; I calculated the price based on a 5 year payback period!
« Last Edit: October 24, 2008, 09:47:00 PM by Zeno »

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #84 on: October 23, 2008, 02:00:53 PM »
Just a quick explanation: I called your previous prices (before re-download) 10-year prices because they resembled those from some earlier Excel. If one replaces in the above posts 10-year by "older version, higher purchase, lower running cost" and 5-year by "the lastest version, lower purchase, higher running cost", it should hopefully clarify it.

Let me also add that higher purchasing price does not necessarily mean tougher game play - purchase is just a lumpsum and after a long time, only running costs and profits determine whether a player makes profit (obvious). So higher running costs are in fact harder to play (provided enough initial capital).
« Last Edit: October 23, 2008, 02:17:55 PM by jatypc »

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #85 on: October 23, 2008, 08:33:22 PM »
I've done another test: this time, I built two transport (coal and waste) with two kristina's + 10 small ore wagons. The results are:
Year193119321933
Income23409,1521486,6222843,82
RC8044,267633,777851,03
Mant.   11116,80   11116,80   11116,80
Benefit4248,092736,053875,99
Payback41,7964,8845,80
So *real* payback periods are quite high I think... :-\
If noone does it before, I'll repeat this test with both the official pak and the modified one; I have no idea of how it was before, but the income trains were gaving seemed *normal* to me: were being paid between 1000 and 1500 cr for a medium run. I'll have to see if costs are also similar than before. But you'll have to wait till next program...  ;D


Edit: Some notation for those who are not familiar with these!
Benefit = Income - RC - Benefit
Payback (in years) = 177520 / Benefit
177520 is the total cost of the two trains, 88760 each.
« Last Edit: October 23, 2008, 08:35:38 PM by Zeno »

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #86 on: October 23, 2008, 09:59:49 PM »
That is what I described in my previous post (the one starting with NOTE) - with the new prices, the payback time in real scenarios can be 10 year if everything is done well, which means among else that trains are transporting without much waiting for the goods. In a similar case like you did I managed get profit with coal trains of 12 waggons roughly between 6-8000 without much waiting and the price of tested trains was certainly below 60-80000 with the latest prices. As I wrote you, I will have a more detailed look and report over weekend.

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #87 on: October 24, 2008, 08:46:35 AM »
To support my claims above and check whether we have similar setups etc., I have done following: run last nightly simutrans + last nightly pak128 with original prices and with your latest prices. The goal was the transport of bulk goods (they pay the least amount per ton and are typically first transported goods) in 1930 with different locos and bulk waggons. Important is that the train was always able to run without any waiting for being loaded 100% (which is not the case if one does not adjust the train capacity to the production). Thus, this is in a sense the optimal case except for maintance - I kept depot etc. unlike AI.

Old prices:

          RVG-2-5-1 + 11      JNC_R11 + 8         RVG-2-4-0 + 12
Purchase  31970               20360               48430
Income    29454               19287               33443
RC        -9445               -5232               -13049
Maintain  -10401              -8421               -10005
Benefit   10795               5633                10388
Payback   2.96                3.61                4.66


Your latest prices:

          RVG-2-5-1 + 11      JNC_R11 + 8         RVG-2-4-0 + 12
Purchase  66630               44540               120710
Income    29454               19287               33443
RC        -9281               -5155               -12288
Maintain  -10401              -8421               -10005
Benefit   9772                5710                11150
Payback   6.81                7.80                10.82


Conclusions:
- new purchasing prices are slightly more than twice higher (this is OK)
- running costs are very similar between both version because RC of locs are a bit smaller and RC of waggons are a bit higher under the new price scheme
- profit is thus approx. equal under both prices, and because of the purchasing price, the payback more than twice the years it used to be (also fine)
- real payback can thus be around 7 year if transport is set up wisely (train runs most of the time, its capacity is appropriate, etc.). In the above example, JNC_R11 accelarates slowly with heavy load, drives less kilometers per year, and thus earns a bit less than RVG-2-5-1. An extreme example is the last column with RVG-2-4-0, where the loc is powerful compared to the other two and can run at more than 100 km/h while waggons allow only 50 km/h - this expensive loc is not a good choice here and one sees it on the payback.

Hence, I think that paybacks around 7 years are feasible under your prices, which is fine with me. Moreover, you achieve this fast payback only with a good choice of transport chain (size of perrons, locs, number of waggons, ...), which is also good.

Finally, the only real change seems to be (at least for the waggons / locs i have checked so far) mainly the purchasing price as running cost etc. are approximately equal. The 2-3 times higher prices thus means 2-3 times higher payback time and nothing else compared to the original prices in the pak128 nightlies.

-----

Additional illustration (approximate, not precise) for a small slow loc with just 4 wagons, but maintanance cost kept to the minimum for such a small train. The results are the same as above and real 7 years payback is achieved


Old prices:
          1Dampflokomotive + 4
Purchase  9380
Income    8761
RC        -2541
Maintain  -3370
Benefit   2760
Payback   3.39

New prices:
          1Dampflokomotive + 4
Purchase  20420
Income    8942
RC        -2553
Maintain  -3370
Benefit   3018
Payback   6.8
« Last Edit: October 24, 2008, 09:41:56 AM by jatypc »

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #88 on: October 24, 2008, 10:09:01 AM »
Additional note: while the above comparison shows things are fine and working as expected in 1930s, the story will be probably quite different in very late stages of the game. Initial checks show that running cost are slightly higher for electric train engines introduced after 2000 (but nothing extreme), and at the same time, purchasing prices can increase sometimes very high, sometimes just a little bit. This is not necessarily bad, one just has to check the extreme cases. Extreme examples with purchase price change out of 2-3 interval:

- 1Benst138: 75900 (2.59/km) ==> 84300 (1.41/km)
- SBB_Re460: 90300 (2.72/km) ==> 328000 (4.63/km)
- RVg_Spike_Duplex (front): 73600 (2.36/km) ==> 482100 (5.36/km)

I do not however know how the coefficients for hydrogen / electric / steam engines are set up and it could be that they initial (old) prices were just inappropriate. The RVg_Spike_Duplex train, mentioned last above, used to be profitable with old prices even if it had just front+end engine + 4 waggons and 50% average load. This will no longer be the case with the new prices and will change the use of such a train (actually with 4 waggons, it can still show profit if running 85%-90% full on average [tested], but the payback will be slow [20-30 years], which does not matter so much given that these trains are never obsolete in the game) .

« Last Edit: October 24, 2008, 11:08:05 AM by jatypc »

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #89 on: October 24, 2008, 01:47:37 PM »
Finally, the only real change seems to be (at least for the waggons / locs i have checked so far) mainly the purchasing price as running cost etc. are approximately equal. The 2-3 times higher prices thus means 2-3 times higher payback time and nothing else compared to the original prices in the pak128 nightlies.
I don't know what do you mean. I changed only cost and RC, and it has turned to be similar RCs than before (some trains change) and higher purchasing costs. I'll try post a list of significant changes as soon as I can, so we'll know what to look at 8)

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #90 on: October 24, 2008, 02:36:54 PM »
I meant that after balancing and change of the cost and RC, RC are similar to those before (just as you say) and it is thus easy to guess what the change in play will be compared to the original prices (this was not the case of the prices originally posted by you if I remember correctly) - this makes life certainly simpler. Moreover, since the purpose of the change was to make prices generate by a rel. simple system, to keep most things similar, and to make it a bit more challenging (higher purchasing prices) - you succeeded so far  :)

Regarding the significant changes, I do not have the current excel sheet and do not see how prices are generated - thus, I can point to the extreme cases and check whether they are profitable or not and to which extent, but I can only guess why they are as they are.

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #91 on: October 24, 2008, 06:27:06 PM »
By now I'm in the middle of a re-make process of the excel sheet. When finished, I will update the link to the new one.
Important question:
I hope the vehicle list I got from simutranslator is the current and valid one, isn't it???
I think so, at least the number of locos is the same... but maybe some were replaced.

Changing to another thing: I've discovered (well, I knew it but didn't realise) there is a time factor affecting to the purchasing cost. It's configured to 0.25% going from 1960 (reference year). But it comes down instead up, I mean, it makes vehicles cheaper as their release date raises. I really think it should be better upside down: as newer locos should be a little more expensive because they're technologically more advanced, and also because of inflation, we all know. It would be a good way to simulate both things togheter, not in a very realistic way but in a very easy way. What do you think about?


Edit: A quick list with heavy changes on vehicles:

The following have running cost arisen by 50-76%:
skoda163, RVg_M1_tail, Haru_E3_KOMACHI_(mid), Haru-BR182_Taurus, 3Bens185, Haru_Spike-PSE_Head2, Haru_Spike-PSE_Tail2, Monatetsu_Steam_Ceres, Haru_E3_KOMACHI_(rear), SBB_Re4_4_II, SBB_Re6_6, SBB_Re460 and 1Bens155.

The following have running cost arisen by more than 94%:
skoda151, Haru_E3_KOMACHI, DB_BR103_1, SBB_IC2k_corch_Bt, SBB_IC2k_corch_Bt_tail, Haru_Spike-PSE_Head1, Haru_Spike-PSE_Tail1, Thunder_(front_engine), Thunder_(rear_engine), RVg_Hiawatha, RVg_Spike_Duplex, RVg_Spike_Duplex_tail, RVg_Daylight, Haru-ACE1_Head and Haru-ACE1_Tail.

There are no vehicles with RC lowered more than 50%. Most lowered are RVg-Pioneer_Alladin and wagons (47% all of them), 1Benst138 by 45.7%, vsys_garatt all three between 40 and 44%...

I'll release a new excel soon, as statistics here on the board are insane, I know ;D

« Last Edit: October 24, 2008, 06:54:55 PM by Zeno »

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #92 on: October 24, 2008, 09:57:09 PM »
I've just built a new version. Some running costs corrected: f.i. Haru ICE1 (6 wagons) had RC about 21cr/km ~= Airbus A380!!!  ;D

Now there is a new version of the vehicles.all.pak. Please, re-download before further testing!!!
Current pricing included in wernieman's nighly downloads!!
« Last Edit: November 26, 2008, 05:16:10 PM by Zeno »

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #93 on: October 27, 2008, 09:19:01 AM »
Changing to another thing: I've discovered (well, I knew it but didn't realise) there is a time factor affecting to the purchasing cost. It's configured to 0.25% going from 1960 (reference year). But it comes down instead up, I mean, it makes vehicles cheaper as their release date raises. I really think it should be better upside down: as newer locos should be a little more expensive because they're technologically more advanced, and also because of inflation, we all know. It would be a good way to simulate both things togheter, not in a very realistic way but in a very easy way. What do you think about?

I agree - one can imagine that a purchase price of a loc 15 years after its introduction might be lower than at the beginning (but that can be implemented only within simutrans exe), but new engines are typically getting more expensive at the time of their introduction due to more advanced technologies. Thus, I think this percentage should have a positive (increasing) effect on the price for railroad vehicles. Considering 0.25% from 1960, the latest new vehicles originate in 2040, which means max. 80*0.25=20% increase - that is acceptable.

On the other hand, I am not sure whether one can use the same argument for the road vehicles / trucks - maybe only for busses, because for trucks i feel the cost grow with better fuel economy and smaller RC, but we do not have this effect in current prices (well, they are still to be defined). So the price increase should probably be modelled differently for good trucks? It is open for discussion as well as prices for all other vehicles.


Edit: A quick list with heavy changes on vehicles:


I have downloaded the latest prices. I will have a look at the profitability of the vehicles with the largest changes in price structure and report "optimal" payback years. For example,

* RVg_Spike_Duplex train of total length 12 running on average 90-95% full on the best track of a relatively short length (it cannot go too fast - max. 220 km/h - and mileage per year is kept reasonable) has in year 2050 with the new cost structure 6.4 years payback (Price ~ 1.6 mil., Profit-RC ~ 325 th., maintanance ~ 75 th.). This looks fine as it does not deviate much the locs tested in 1930s.

« Last Edit: October 27, 2008, 09:36:31 AM by jatypc »

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #94 on: October 27, 2008, 09:43:30 AM »
Mmm... That sounds even better than I expected  8)

I've had an out-home weekend, so I have not been able to test at all  :(
I'll try to get back to it during this weekend (despite having a lot of work).

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #95 on: October 27, 2008, 09:51:59 AM »
I have downloaded the latest prices and post results by updating this message.

I however wonder whether I have right prices (or where else an error could be) because while RVG_Spike_duplex looks reasonable (e.g., cost=408200/RC=3.11 for the head), the prices for Haru_E3_Komachi look strange. The Haru_E3_Komachi set consists of always of 4 waggons: head (1200 kW, cost=12400, RC=0.11), mid (1200 kW, cost=202900, RC=1.68), trailer (just passangers, cost=44420, RC=0.53), rear (1200 kW, cost=30400, RC=0.25). This results in very low cost on front and rear engines and very low RC altogether, but you reported this engine as the one with increased cost. At the same time, this train does not make all that much profit so it is not like it could bear high RC cost.

* RVg_Spike_Duplex train of total length 12 running on average 90-95% full on the best track of a relatively short length (it cannot go too fast - max. 220 km/h - and mileage per year is kept reasonable) has in year 2050 with the new cost structure 6.4 years payback (Price ~ 1.6 mil., Profit-RC ~ 325 th., maintanance ~ 75 th.). This looks fine as it does not deviate much the locs tested in 1930s.

* Thunder train set up in the same way as RV_Spike_Duplex above (although 12 is too short for this train), 90% full on average in 2050 etc. resulted in 7 years payback despite rel. short length and very high purchasing cost (2.5 mil.) because it accelarates and goes faster.

* RVg_thunder_2: compared to Thunder_(front engine) and Thunder_(rear_engine) in the previous point, it is cheaper (440 th. instead of 791 th.), it is more powerful (7700 kW instead of 4900 kW per engine), it is faster (370 km/h instead of 250 km/h) and it has lower RC (13.54 instead of 18.53 per train of length 12). The question is why - why some much cheaper and lower RC when it is more powerful and faster?

* RVg_M1: again the same setup as before with trains of length 12, etc. Here one sees a strong effect of speed bonus. At 2010, the payback in years would 4.5 (train of length 12 earning 175000 per year). The number is OK I think because it looks like a regional train, so in practice trains would be shorter than 12 (although the train runs fine also at this length). In year 2050, the train running almost always 100% full earns about 75000 per year after subtracting the running cost. Because it costs about 550000, its payback would be 10-15 year depending upon how many trains share the maintanance cost.

* to be re-tested, payback years will go up SBB_Re460 with 11 waggons of the same speed (Marks_passagierwagen), year 2050, running 90% full as in previous test, payback in year is 4.7 because of relatively small RC of the train (but that comes from waggons as the standard passanger train cars have lower RC than the special train cars used in high speed trains).

* to be re-tested, RC increased a lot Haru_ACE1 train of length 12, year 2010 (20 years after its introduction, 10 year before being obsolete) with the setup as before has payback about 6.7 years (and smaller after its introduction) (yearly earnings after RC and maintanance subtracted ~ 300 ths., purchase cost ~ 2 mil.)

* Skoda151 with 11 72p-waggons of the same speed in year 2001 results in payback period around 7 years, one can however use also 100p-waggons without slowing the loc down or increasing the weight of the train, which would lead to a lower payback period.

* Skoda151 with 11 72p-waggons of the same speed in year 2001 results in payback period around 7 years, one can however use also 100p-waggons without slowing the loc down or increasing the weight of the train, which would lead to a lower payback period.

* DB_BR103_1 with 11 72p-waggons of the same speed in year 2001 results in payback period of 6 years (it is expensive, but also faster that Skoda151, so it gets higher speed bonus).

* RVg_Tigress of length 12, year 2050, price ~ 2.6 mil., profit per year - RC - maintanance ~ 430 ths., payback years ~ 6 years

* JR_700_Series_Shikansen_Hikari_Rail_Star of length 11 (longer waggons) in 2050 (tried because of different structure - every waggon has an engine), price ~ 2.7 mil., profit per year - RC - maintanance ~ 230 ths., payback years ~ 11.7 years; the same in year 2010 (to check speed bonus effect): price ~ 2.7 mil., profit per year - RC - maintanance ~ 320 ths., payback years ~ 8.5 years. This is generally more than for the trains consisting of fixed number of engines and waggons without engines.
« Last Edit: October 30, 2008, 09:11:23 AM by jatypc »

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #96 on: October 27, 2008, 11:07:27 AM »
Great feedback! Thanks jatypc!!

Let me comment a couple of things:

i) RVg_thunder_2: Let me take a look on it. I'll see what can I find out when I'm back home tonight.
ii) RVg_M1: Actually is what I expected. I pretend to make all trains have similar conditions, so all RCs and prices calculated same way, but I don't expect bonus system and time advance to affect all they in the same way; I think that's not our war now ;)
iii) Others: I'm quite happy overall tests seem to be keeping playability while extreme cases (big RC/price changes) seems to keep some logic and are integrated. That is our goal, ain't? ;D

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #97 on: October 27, 2008, 11:16:50 AM »
Yes, I agree the bonus should not be our concern - if a train is purchased when it is relatively new, it earns money easily; in later years, it has already earned its cost, so it just has to cover RC and maintanance and give some extra profit.

Quote
I'm quite happy overall tests seem to be keeping playability while extreme cases (big RC/price changes) seems to keep some logic and are integrated. That is our goal, ain't?

Yes, I hope I will be able to see some logic in the final Excel sheet too :P (you are by now much more familiar with it  ;))

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #98 on: October 27, 2008, 11:44:59 AM »
Oh %@&¿$h&!!! I forgot to update the excel link!  :-[

Yes, I hope I will be able to see some logic in the final Excel sheet too :P (you are by now much more familiar with it  ;))
Don't expect to see any logic there... it's more like a mental/ability game between battleships and sudoku ;D ;D ;D

Offline VS

  • Senior Plumber (Devotee)
  • Moderator
  • *
  • Posts: 4852
  • Vladimír Slávik
    • VS's Simutrans site
  • Languages: CS,EN
Re: PAK128 Pricing & Balancing - Discussion
« Reply #99 on: October 27, 2008, 04:39:48 PM »
Looking good so far.

The MUs can be balanced separately if you need… I guess there aren't that many?

My projects... Tools for messing with Simutrans graphics. Graphic archive - templates and some other stuff for painters. Development logs for most recent information on what is going on. And of course pak128!

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #100 on: October 27, 2008, 10:06:44 PM »
As I promised, I leave here the link to the last excel version I have been playing with. I've made lots of small updates here and there. Biggest changes are in speed_bonus sheet, where I tried to reflect actual way of doing things with a list of years and speeds (like the tab file, but calculated for every year). Also some big changes in the engines (locos) sheet: looks basically the same, but some formulas have been adapted. Also I did a small remake version of the macro: now it takes 30s in my computer (I disabled screen updating and removed cell selections to use cell values). Anyway, questions and doubts are addmited ;D

The MUs can be balanced separately if you need… I guess there aren't that many?
That's a possibility. Let's see what differences we find between "non-cargo-engines" and MUs and then decide. But my opinion before looking at it is to balance separately if they don't fit in current measures. But by now, first tests are looking quite nice for me: cargo engines are quite less sensible to losses (as they are calculated to run back empty); meanwhile MUs (I've tested a couple EMUs) are quite profitable, but if they get loaded below 80% they can sink your company :o And I like that way ;)

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #101 on: October 28, 2008, 09:52:21 AM »
Thanks for the Excel file - it clarifies some things. Most importantly, strange prices and cost for Haru_E3_Komachi and some other vehicles are explained by the difference between the Excel prices in tabs Engines & Waggons and those in tab Export. For example, the prices for Haru_E3_Komachi_(rear) are 30400/0.25 in simutrans and Export tab, but they are 215000/2.66 in tab engines (there are more examples of this). Is this intended? Without knowing what the correct prices are it is difficult to test.

By the way, Haru_E3_Komachi can never be profitable with the current prices in tab Engines, which comes from the fact that there are 3 (passanger carrying) engines in that train and only one waggon, but the entries in the Excel sheet look like each engine pulls 3 waggons. Putting the number of waggons to 1 should give more or less reasonable results I guess.
« Last Edit: October 28, 2008, 09:56:37 AM by jatypc »

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #102 on: October 28, 2008, 10:13:10 AM »
Mmmm... The export sheet is just a snapshot in a certain time; you should take the engines one as reference, but beware the values in red cells are the correct ones. They're supposed to be 15% for income/rc, 100% for bonus influence. The other ones are just calculations. After adjusting them remember to recalculate effective speeds using the macro (click on calculate eff_spd button). Then you should get correct values in the engines sheet. Tell me if you have any trouble with that.

I'm thinking we must calculate MUs having fixed wagons number separately... The Komachi is one of them, isn't it?
« Last Edit: October 28, 2008, 10:23:33 AM by Zeno »

Offline jatypc

  • *
  • Posts: 49
Re: PAK128 Pricing & Balancing - Discussion
« Reply #103 on: October 28, 2008, 10:54:57 AM »
Mmmm... The export sheet is just a snapshot in a certain time; you should take the engines one as reference, but

OK, then the "vehicles.all.pak" does not correspond to the current Excel and I will try to test only vehicles, where I have the same numbers...

Quote
I'm thinking we must calculate MUs having fixed wagons number separately... The Komachi is one of them, isn't it?

Yes, Haru_E3_Komachi, Haru-IC3, RVg_flying_berliner, Haru-373, RVg-Pioneer_Alladin, Evil_BR357, Evil_BR170, 471, nankia21K, Haru-PlanV have all fixed length larger than 1.

How do the units (rail busses such as DC22) consisting of one engine carrying passagers and no waggons fit in the picture? (Haru_VT98, Dc22, JRW_Mc123, Evil_BR153)

By the way, it seems that high speed trains with a fixed number of engines and waggons earn back their purchasing price generally faster than the high speed trains consisting of engine-waggons (each waggon has an engine such as Shikansen trains).
« Last Edit: October 30, 2008, 09:16:53 AM by jatypc »

Offline Zeno

  • ENASSA Designer
  • Devotee
  • *
  • Posts: 1997
    • Zeno's Simutrans Creations
  • Languages: ES, EN, CAT
Re: PAK128 Pricing & Balancing - Discussion
« Reply #104 on: November 10, 2008, 08:21:23 AM »
I'm sorry I've been quite busy these days, so I've been absolutely disconnected. I'll try to get back to work after finishing some homework during this week :-[