Welcome, Guest
Username: Password: Remember me

TOPIC:

Benchmarks 7 years 10 months ago #5838

One thing I think would be great would be some kind of benchmark system so you know how fast you are on a server - the fastest times lists set a target for the best players, but from my experience (unless I'm missing some neat trick) these are only achievable under specific conditions like starting the game 5-10 years in. It would be great to have some way of knowing what a good time is from a standing start.

Suggestions that might work (for each server):
- A "fastest times" that only includes players starting in the 1st year
- A graph showing %age of winning times (eg top 10% finished in under 1hr, 20% in under 1:10 etc...).. and again perhaps limited to 1st years starts.
- A static rating guide (eg "beginner: 2hrs, average: 1:30, expert 1:15"), perhaps even just a forum post

This would give an idea of how much room for improvement you have in your game.

The main reason I'd like this is because I've hit a wall of about 1:10-1:15 in the 25t CV server #3, well outside the fastest times (from last year, I'm #6 so far this year at 1:19! Last years scores were all <45min) and I'd like to know if my game play could be improved, or are faster times just people starting late.

If anyone knows what the fastest from-day-1-times are for Server #3 please tell me, but a table/graph/guide for each server would still be great because I'll play others servers in the future and will be in the same boat.

If it's simply a matter of crunching the data from last year I'm happy to do it if you send me the data.
  • Tobus
  • Tobus's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 38
  • Thank you received: 19

Please Log in or Create an account to join the conversation.

Benchmarks 7 years 10 months ago #5839

Hi Tobus,

I think it's a great idea, but unfortunately my knowledge on these subjects is next to 0 (that is zero ;-)).

I could share with you the data of last year in the form of a MySQL data dump (I can also share the user base, but without email addresses and Password fields of course), cause our database is relative to userid in the most cases.

Let's talk on IRC if you have some time, the connection details are under the contact us page... ;-)

Thanks for your suggestion though, I think it would be cool indeed if we make it happen!

Frank
Webmaster and Server Admin BTPro

Please Log in or Create an account to join the conversation.

Benchmarks 7 years 10 months ago #5842

Hi Tobus!

During last year several configuration updates were made to make the game more interesting/fair for the players, e.g. APHID CBs were changed (#27-#30) to the newer script, making the game slightly more challenging, thus we were unable to reach the best times set in the first half of the year. I'm sure that config of CV servers #5 and #17 have been also changed, I don't know #3 so well.

Taking above into account + starting year + map randomization = I cannot imagine a reliable way for static rating guide.

Anyway, I like the idea of fastest times including only the games starting in the 1st year. But it should be optional imho (e.g. checkbox to show fastest times like they are now/new style). Number of competing companies would also be nice (excluding companies created automatically on joining the game and removed one month later, because the player left the game).
  • tomb
  • tomb's Avatar
  • Offline
  • -=EPIC Donator=-
  • -=EPIC Donator=-
  • Posts: 237
  • Thank you received: 234

Please Log in or Create an account to join the conversation.

Benchmarks 7 years 10 months ago #5844

Frank sent me dumps of all the game data - unfortunately there's no "starting year" or similar field so it's impossible to make a "from day 1" list. It's something they've flagged as possibly including in the future, so maybe next year.

I've gone through the data and come up with the following table:

Srv Average Good Expert
#1 3:01 2:25 1:50*
#2 8:51 7:04 6:06
#3 1:43 1:20 1:06
#4 3:35 2:46 2:19
#5 6:07 4:50 3:53
#6 1:53 1:30 1:16
#7 2:25 1:59 1:45
#8 1:53 1:30 1:17
#9 5:39 4:34 3:55
#10 2:36 1:50 1:29
#11 4:06 2:30 1:56
#12 1:01 0:51 0:45
#13 0:50 0:40 0:35*
#14 2:08 1:38 1:25
#15 1:52 1:25 1:15
#17 12:53 10:35 9:16
#18 3:24 2:40 2:23
#19 3:16 2:37 2:16
#24 1:05 0:54 0:47
#25 0:50 0:40 0:35*
#26 1:45 1:22 1:12
#27 2:20 1:32 1:14
#28 2:48 1:59 1:38
#29 2:16 1:23 1:15*
#30 3:41 2:27 1:58
#31 2:05 1:39 1:26
#33 1:59 1:40 1:30

* - manually edited as per comments below


These represent roughly the fastest 50%, 25% and 12% of games on each server, after extremely slow/fast game were removed - not perfect, but should give an decent target to aim for. Would be great if you and other experienced players can look over an see if anything looks odd.

I note that #12/13 and #24/25, the 4k/8k CB servers, are very similar but from my experience the 4k server takes about 5-10min less. Perhaps beginners start on the 4k and when they start getting the hang of it move to the 8k and all their fastest games are there? At any rate, I'd suggest the following change:

#13/#25 0:55 0:45 0:40 (now updated in table)
The following user(s) said Thank You: Frank
  • Tobus
  • Tobus's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 38
  • Thank you received: 19

Please Log in or Create an account to join the conversation.

Last edit: by Tobus.

Benchmarks 7 years 10 months ago #5845

The fastest 15 games should simply be reset to empty list at each configuration change for a server. In such way the list will provide useful information for players.

Please Log in or Create an account to join the conversation.

Benchmarks 7 years 10 months ago #5846

Nice work, Tobus! And now I'm sure that the results would be much better if at least the following data is taken into account:
- starting year: skip games not started in the 1st year,
- number of players in company: skip results obtained by team of players (these games are usually finished faster).

The most odd results for me:
#1 expert mode: 1:50 would be better, I've seen a few players last year who easily went below 2 hours starting 1st day,
#29 expert mode: 1:06 is a matter of luck/late year, 1:15 is hard to get,
#9: compare your results to this year fastest games, please.

Other thoughts:
- where are results for servers #31-#33?
- what about servers with different climates (#33, #9)?
  • tomb
  • tomb's Avatar
  • Offline
  • -=EPIC Donator=-
  • -=EPIC Donator=-
  • Posts: 237
  • Thank you received: 234

Please Log in or Create an account to join the conversation.

Time to create page: 0.071 seconds
Best hosting deal on hostgator coupon or play poker on party poker
Copyright 2020 Benchmarks - BTPro - OpenTTD Community.