News:

Use the "Forum Search"
It may help you to find anything in the forum ;).

PAK128 Pricing & Balancing - Discussion

Started by DirrrtyDirk, October 10, 2008, 03:21:44 PM

Previous topic - Next topic

0 Members and 2 Guests are viewing this topic.

VS

Yes, it is a possibility.

However I would like to use also some brain cells, not brute-force ;) If you take a look at the trends of average and median, they are more or less "stairs" or linear bits.

So let's combine percentages with rough interpolation by human factor. 75 is good guess - comparing the values and what I saw while playing, it could be a bit too generous, but... let's see.

More to follow tomorrow.

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!

Zeno

Yeah, some combination could be applied to get best results. F.i., I'm playing a game and just detected on 1965, my boeing 707 were earning tons of money, about 700-900K per year. Later on, about 1972, maybe 1973... it gets losses!!! I think that could be calibrated better, as long as there are only 3 or 4 values for planes. Trains & road vehicles won't be so easy to put in place  ::)

jamespetts

I know that I have come to this discussion somewhat late, but I should like to add some general thoughts on balancing philosophy if I may. Firstly, in my view, it is a mistake to have all vehicles have exactly the same value for money (having exactly the same ratio of purchase price to running costs to haulage capacity to maximum speed for each cargo type) because it makes the choice between different kinds of vehicles less interesting (and also less realistic). Some vehicles would be better value for money than others, and some, although worse value for money overall, would be more profitable in certain specialist situations (such as crack express trains, or short-haul bulk transport at slow speed, etc.). There should perhaps be an element of diminishing returns, in that paying ever more for a vehicle makes it better in increasingly small increments, the player having the difficult but interesting decision in every case where to stop. (The ideal manifestation of this would be to have several different price bands of vehicles all available at the same time, and then, as technology moves on, those same price bands applied to respectively higher specifications). There should also be a certain element of randomness: some vehicles should just be plain better than others (not necessarily be large amounts, but enough to make a difference). Again, this makes gameplay more varied and interesting, especially if players cannot rely on improvements in technology to be consistent.

Also, the game would be more interesting if the ratios changed over time. The trend in reality has been (very approximately) for newer vehicles to have higher purchase prices, but lower running costs and/or better performance. If this pattern was implemented in the game, the result would be a game in which both the risks and rewards slowly increased with the passage of time, starting out with it being fairly easy to make a small profit, but difficult to make larger margins, but becoming increasingly the case that, if one has the money to invest, and one manages the network well, one can make very considerable margins on newer vehicles, but there is a far higher risk of overspending and losing a very large amount of money quickly if one makes a mistake. That again would make the game more interesting.

Furthermore, I have been working on some patches to the main code which, if accepted, could make a real difference to how the paksets should be balanced. Further discussion can be found, amongst other places, here, which also contains an initial version of the first half of the patch. (I am still working through some techincal problems on the second half). In essence, the function of the patches would be (1) to make the speed bonus (either positive or negative) have less and less impact the shorter the journey; and (2) gradually increase the maintenance costs of obsolete vehicles from the date of their obsolescence until (by default) 20 years after retirement. The cumulative effect of those changes would be to make differently balanced vehicles preferable for short-haul transport than are preferable for long-haul transport. There should be high capacity vehicles with low running costs but a low speed for short haul transport, but lower capacity vehicles with a higher top speed and higher running costs for long haul transport, since the earnings with short-haul transport would not vary much with speed, and so would benefit from capacity and low cost instead. These modifications would also mean that the speed bonus ought be set perhaps a little more aggressively than it is now, so that the speed that gets the 0% bonus is faster than the speed of the vehicles designed just for local transport (and is, perhaps, the approximate average of older but not obsolete vehicles designed for medium distance transport).

Finally, may I ask - are the British and 19th century vehicles available on addons.simutrans.com part of this great rebalancing? I always like playing with those vehicles, and it would be a great pity if they were not included.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

DirrrtyDirk

Yes, you are late. Very. I don't know how far along the work actually is right now, but we are trying to get a new official release of pak128 out as soon as we can manage. And my guess is that redoing everything from the start (once more!) isn't helping us to get there... So my guess is that these ideas will probably not get into the next release. After that, who knows... but let's walk step by step - first things first.

Oh and all vehicles using the same ratio of value for money might not be realistic - but it a) makes things easier for new players (no danger of buying a "wrong" vehicle, since they are all comparable and calculated by the same method) and b) it allows for more diversity (otherwise you'll probably end up with people only using the single best vehicle, instead of allowing them to choose one from several equals... at which point we could almost remove 80% of the not so efficient vehicles completely, because nobody would be using them anymore). So... no, I don't think that's a good idea - not for the near future at least. We are actually trying to get things simpler and easier to understand, not more realsitic (and therefore also more complicated).

And since there is a whole pakset for Britain and British vehicles of itself, I don't think any additional ones are needed in pak128 at this point. Also, once again: there's a reason for addons to exist. Not everything that is created - and liked by a couple of people - needs to be part of the basic sets. So if you like these vehicles and they are available as addons - just use them.  ;)
  
***** PAK128 Dev Team - semi-retired*****

jamespetts

Quote from: DirrrtyDirk on December 11, 2008, 12:00:14 AM
Yes, you are late. Very. I don't know how far along the work actually is right now, but we are trying to get a new official release of pak128 out as soon as we can manage. And my guess is that redoing everything from the start (once more!) isn't helping us to get there... So my guess is that these ideas will probably not get into the next release. After that, who knows... but let's walk step by step - first things first.

Ah. Note to self: join in earlier next time ;-) Of course, once the new basis for calculation is set up and all the vehicles have their figures generated on a standard basis, it would be relatively easy to make small adjustments based on the suggestions here.

QuoteOh and all vehicles using the same ratio of value for money might not be realistic - but it a) makes things easier for new players (no danger of buying a "wrong" vehicle, since they are all comparable and calculated by the same method) and b) it allows for more diversity (otherwise you'll probably end up with people only using the single best vehicle, instead of allowing them to choose one from several equals... at which point we could almost remove 80% of the not so efficient vehicles completely, because nobody would be using them anymore). So... no, I don't think that's a good idea - not for the near future at least. We are actually trying to get things simpler and easier to understand, not more realsitic (and therefore also more complicated).

I don't agree with this, for two reasons: (1) because making things as simple as possible is not the way to make a game fun; and, more importantly, (2) because I was not suggesting that any vehicle be balanced so as to be the "wrong" vehicle overall: every vehicle should be balanced such that it is the right vehicle for at least some situations, but, just as in real life, some vehicles will have a wider range of situations for which they are the right vehicles than others. A vehicle that is, overall, worse value for money than another vehicle can still be the best vehicle to use in certain situations. It might also be that, as technology advances, vehicles gradually become better value for money in some respects, and worse in other respects. Aircraft might well become better value for money as time progresses, for instance, and trains worse value for money.

As to the British pakset - do you know whether that was ever finished?

And since there is a whole pakset for Britain and British vehicles of itself, I don't think any additional ones are needed in pak128 at this point. Also, once again: there's a reason for addons to exist. Not everything that is created - and liked by a couple of people - needs to be part of the basic sets. So if you like these vehicles and they are available as addons - just use them.  ;)
[/quote]
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

jatypc

james, let me add that it is true we are trying to make profitability similar across vehicles. On the other hand, the rules to set up pricing are relatively simple, and therefore, there is quite some variability in profitability among different vehicles (e.g., years to earn the purchasing costs can vary, for example, from 4 to 9 in more or less "optimal" setup - make something wrong and profitability is gone). Moreover, there are many other factors influencing the choice of a vehicle:

a) speed and capacity (small vs. large number of vehicles, sufficient vs. insufficient service)
b) speed vs. the cost of infrastructure
c) number of convoys vs. the structure of infrastructure (connections)

To choose the optimal vehicle for given route is thus far from trivial and variability as well as chance of having money-losing vehicle is there...

VS

I can't really say anything about Zeno, but I just cower in the corner and hope it goes away soon :D

A few minutes ago, in r275, I set the speed bonus to values that roughly match average speed. Of course it is interpolation of uncertain data by a set of linear functions, so there are quite a few rough jumps. I tried to keep avg(speed, time) >= speedbonus(time). This means there are some fat and some hungry times - depending on how big the actual difference is. Btw, the old values were complete nonsense, sometimes completely out of the min/max bounds...

Monorails should now make some money - previous speed for zero bonus was in the lowest place 20% higher than maximum.

Ships also had these gross extremes, so now the trend is more sane. The new values are set so that average speed is almost identical, except for a plentiful period in 1900-1950.

There is a drop in tramway speed after 1910, so trams should be very profitable before 1910... if you can get any!

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!

prissi

Why was everyone ignoring the speedbonus.tab I had exactly made for pak128? VS, you could have used it ...

VS

Oops, I didn't see it there. The new values are very similar indeed. In my defence, I wasn't the only one to overlook it :(

At least I learned something about the speeds - and printed them on paper...

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!

Izzy

I am trying to do some development and I've got problem with vehicle data. Is there any template or calculator to set vehicle data according to the new balancing rules?
And another question.
How the weight of vehicles is calculated? Because according to official pack Icarus 260 weight 28 tons! witch got nothing to do with real life. Is that weight of bus with full passengers load or is it only vehicle?

Cheers   

jamespetts

The weight should be the unladen weight, since the program adds the weight of the freight dynamically.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Zeno

#151
Quote from: Izzy on January 06, 2009, 10:11:50 PM
How the weight of vehicles is calculated? Because according to official pack Icarus 260 weight 28 tons! witch got nothing to do with real life. Is that weight of bus with full passengers load or is it only vehicle?
Cheers   
The balancing only changes Running Cost and Vehicle Cost. Vehicles weight is not calculated but coded in the PAK.
I use the value the creator put there, so if Icarus 260 weights 28 tons, I will adjust its RC and purchasing price to match that value.
Btw, if you detect some more incorrect vehicle's data, I would suggest you create a post collecting data issues so they can be modified.


Edit:
AFAIK Icarus 260 and 280 both have (yes, I mean in game!) a weight of 31 tonnes (empty) ;)

VS

Are you referring to reality or the vehicles in game?

It should be 12t...
http://www.busnumber7.com/index.php?id=1

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!

Zeno

Quote from: VS on January 07, 2009, 03:51:20 PM
Are you referring to reality or the vehicles in game?
To game data, of course!! That's why I suggested to create a post with errors in data files :)

VS

Oh! I have felt that quite a lot of road vehicles is somehow weaker than they used to. Some could use a lower weight or higher power... but a systematic approach would be better than guesswork.

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!

Zeno

Quote from: VS on January 07, 2009, 05:07:38 PM
Some could use a lower weight or higher power... but a systematic approach would be better than guesswork.
I can give you a general idea by getting those vehicles which have lowest values for theorical/potential max speed (w/o limit) minus real max speed; that would give us a list of potentially weak vehicles. ;)

emaxectranspoorte

#156
Quote from: VS on January 07, 2009, 05:07:38 PM
Oh! I have felt that quite a lot of road vehicles is somehow weaker than they used to. Some could use a lower weight or higher power... but a systematic approach would be better than guesswork.

Ok, VS. :) Have you heard the joke about the sound of cars, related to their power ... ? ;) ;D I guess you did... :)

If you don't know ... :) I'll PM it to you, as it might upset someone, here... :( :-\

Izzy

I have just discoverd that on latest nightly build pack version "simupak128-325" some buses are still cheaper then others.
KC-B10M Keisei Bus seems to be cheaper then any other bus.
I'm not sure is it after rebalancing version pack but links abowe to test versions are not working.

Cheers 

Zeno

Yes, you may find newer buses seem to be cheaper, but that's the effect of the bonus-speed: newer vehicles need higher speed to get same money. So they're a bit cheaper.

Zeno

#159
I'm doing new testing with trains: I've changed weighting of engine types. I've seen electric trains RC is calculated using same formula and weighting than diesel ones, so if mainainance of electric overheads is a plus on electric locos. That way electric locos are much less profitable, so I'm trying to level it a little.

Original weighting for loco's RC was:

  • RC*1.1 for Steam, RC*1.0 for Diesel, and RC*1.0 for Electric.
  • Now testing with 1.15 steam (+5%), 1.05 diesel (+5%) and 0.95 electric (-5%).
  • I've also raised electric loco's purchase price by 10%.

I'll report results and impressions... That's auto-feedback! :D

Btw, what do you think about the idea? Any suggestions?

jatypc

I think it is a good idea (originally - in 1.00/128 - the eletric engines had smaller running costs than diesels, mostly). My guess it even corresponds to reality as long as one assumes that an electric power plant generates electricity more efficiently or cheaper (by 30-40%) than the diesel engine in dieselelectric locs.

jamespetts

In reality, electrics are considerably cheaper to run and build than diesels, since they are a great deal simpler and more efficient. That is offset to some extent, however, by the infrastructure cost for the delivery of electricity.
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Zeno

Electric locos are cheaper to build than diesel ones? I've always thought they were more expensive...  ::)

jamespetts

Quote from: Zeno on January 13, 2009, 09:52:44 AM
Electric locos are cheaper to build than diesel ones? I've always thought they were more expensive...  ::)

Hmm, why would they be more expensive? A diesel-electric has to have everything that an electric locomotive has, plus a diesel engine...
Download Simutrans-Extended.

Want to help with development? See here for things to do for coding, and here for information on how to make graphics/objects.

Follow Simutrans-Extended on Facebook.

Zeno

Quote from: jamespetts on January 13, 2009, 09:55:34 AM
Hmm, why would they be more expensive? A diesel-electric has to have everything that an electric locomotive has, plus a diesel engine...
Mmmm... yeah, sounds easy ;)
I may take away that price rising...

Zeno

Well... We're entering last stage!
I've been doing some numbers for planes and ships... Actually it's applying same idea used to get trains & road vehicles' values. I've but introduced a new concept, which are some price factors, that raise purchase prices of planes and ships so it gets more like IRL, where a plane may cost 100 to 300 M$ while a train may cost 6 to 10 M$. I've used a weight factor combined with a capacity (in tonns) factor to increase/decrease the pre-calculated purchase price.

These are some global numbers we can get from new values ("." is thousand sep, "," is decimal sep):

  • Ships average cost raised from 47.479,38 to 67.209,07 cr
  • Ships average RC lowered from 14,28 to 10,84 cr/km
  • Planes average cost raised from  319,644.40   to 874.888,62 cr
  • Planes average RC raised from 6,30 to 8,75 cr/km

That's just to get an idea of how modifications affect general/global gameplay. It should be analised at different ages (1930s, 50s, 70s, 90s, fi).

I'll now build a new beta pak so these changes are testeable; also last train modifications will be included: +3% RC diesel locos, -4% RC electric locos. Until then (maybe this afternoon), suggestions/comments will be welcome :)

PS: Also I've introduced a new factor for RC in planes, which is as easy as RC = RC * times_sound_speed_crossed, so planes below 1 match don't get any modification, but planes quicker than sound speed get higher running costs. At the moment only my Concorde is affected, and isn't in the official pak, so no problem with that.

wernieman

You now, that Monday and Friday the Nightly-Computer build a new Pak (if there new Data)?

For This:
If you have check in new data to the svn, give me a mail and I start a special run ...
I hope you understand my English

emaxectranspoorte

#167
Quote from: Zeno on January 16, 2009, 11:43:58 AM
That's just to get an idea of how modifications affect general/global gameplay. It should be analised at different ages (1930s, 50s, 70s, 90s, fi).
That is all related to the profit and/or revenue made in the game, right? :-\

Zeno

Quote from: emaxectranspoorte on January 16, 2009, 12:50:57 PM
That is all related to the profit and/or revenue made in the game, right? :-\
Well, that's related as long as running costs change a bit, so profit will also change a bit (revenue is what is). Also purchasing prices change, so time elapsed until the price gets payed back also changes. But all it will depend a lot on the vehicle and year.

As always, all prices/rc are calculated based on vehicle intro year bonus speed versus vehicle speed. So the vehicles are all balanced respect their intro date. As time advances, all vehicles keep getting a little less competitive, as long as the bonus speed is raising.

Quote from: wernieman on January 16, 2009, 12:48:25 PM
You now, that Monday and Friday the Nightly-Computer build a new Pak (if there new Data)?
For This: If you have check in new data to the svn, give me a mail and I start a special run ...
Thanks Werner, I may upload data today if I feel it's ready; if I think it worths to make a special run I will upload and tell you. Thank you.

wernieman

P.S. I am Freiday and Sunday on the "Autobahn", so the new Version will go tomorow, when you check in today  ;D

So you have time  :police:
I hope you understand my English

Zeno

#170
Quote from: wernieman on January 16, 2009, 08:35:14 PM
...so the new Version will go tomorow, when you check in today...
I've the balance almost ready... Maybe I'll upload the files before going to sleep. I'll load a saved game and let it play for some years (accelerated, of course). I'll see results then if everything ok I'll upload so tomorrow there is new nightly :)


Edit: Just uploaded changes. We'll see if feeback increases now with next nighty ::)

wernieman

There should be a nw PAK128-Version on the nightly-Webpage ....
I hope you understand my English

Zeno

#172
As I think current results are quite acceptable, I'll focus my efforts on migrating current excel sheets to open office calc format. I may do some changes to the format that Tomas K. gave to this sheet, but essentially will keep most of its data and functionality. When finished, I will upload it to svn server.

wernieman

Did somebody hear somethink about the new pricing in the pak128?

positiv or negativ?

The Version is 1 week old ...
I hope you understand my English

Zeno

Quote from: wernieman on January 25, 2009, 09:05:00 PM
Did somebody hear somethink about the new pricing in the pak128?
positiv or negativ?
The Version is 1 week old ...
No news on that... seems the balancing has perfect results ;)
As some other people suggested, I think with the next release will come more feedback... I hope!
Anyway there's still some work to do, specially with troleybuses and trams. So we'll see.