Recommend
3 
 Thumb up
 Hide
13 Posts

Race for the Galaxy» Forums » General

Subject: RftG on Genie - My first 101 games rss

Your Tags: Add tags
Popular Tags: [View All]
Simon Harris
United Kingdom
Sutton
Surrey
flag msg tools
Avatar
mbmbmbmbmb
I've been playing the RftG implementation on the wonderful Genie server for a while now and definitely reached a bit of a plateau. Because I don't understand ELO ratings (I understand the *concept* just not the mathematics!) I've been looking at my performance (for what of a better word!) in a slightly simpler way using MS Excel.

First I cut and pasted out my 'finished games' list (the 'last action' column is irrelevant in this case) into a blank workbook and reverse sorted the data just because Excel likes time going *down* the page i.e. the newest event is at the bottom of the list. Then I made a couple of extra columns to generate my personal rating.

matches
Total number of 'match ups' for this game (i.e. number of players - 1). Use COUNTA to return the number of non-blank cells across the range of 4 possible player score columns.
score
My score for this game. You probably *can* substring the score immediately following your username automatically but that's a bit too advanced for me so I fill in this one manually.
t_matches
Running total of matches (= 'previous' t_matches + matches for 'this' game
t_score
Running total of score (= 'previous' t_score + score for 'this' game)
win%
Running winning percentage (= t_score / t_matches)

Obviously, the advantage of using formulae (rather than manually entered figures) is you just have to paste in new game data and copy down the fields above.

Finally, I charted the sorry story of my personal progress amongst the RftG greats.


My win%, which is rather pathetic, has stayed within a 1% band for the last 20 or so games. I can post the actual data file if anyone is interested but the formulae above are fairly standard Excel fare.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Serge Levert
Canada
Vancouver
British Columbia
flag msg tools
badge
Avatar
mbmbmbmbmb
Pretty neat. You've inspired me to generate some graphs on my own career data, it's about time i did. I have lots of numbers but hadn't thought to make graphs.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Simon Harris
United Kingdom
Sutton
Surrey
flag msg tools
Avatar
mbmbmbmbmb
entranced wrote:
Pretty neat. You've inspired me to generate some graphs on my own career data, it's about time i did. I have lots of numbers but hadn't thought to make graphs.

Now don't get me wrong - I won't do this for everyone - but since your name rang a bell (you played in *my* first game), and to prove how simple this is, I've pulled off your stats as well. Assuming that is, your 'nik' is the same on Genie as on BGG.


Simon
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Serge Levert
Canada
Vancouver
British Columbia
flag msg tools
badge
Avatar
mbmbmbmbmb
Heh thanks!
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matt N

Pennsylvania
msg tools
Avatar
mbmbmbmbmb
I do like the formulas and graph, but from the thread title I thought you'd be asking for advice or something. So... are you asking for advice?

Also, win % is pretty misleading. If you farm a bunch of scrubs you could probably break 70%, but it won't mean that you're good; likewise, if you play only against top ten players, 35% is very respectable for the large majority of players, and means you're well above average yourself.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Simon Harris
United Kingdom
Sutton
Surrey
flag msg tools
Avatar
mbmbmbmbmb
Stunna wrote:
I do like the formulas and graph, but from the thread title I thought you'd be asking for advice or something. So... are you asking for advice?

Not really, I'm pretty philosophical.

Stunna wrote:
Also, win % is pretty misleading. If you farm a bunch of scrubs you could probably break 70%, but it won't mean that you're good; likewise, if you play only against top ten players, 35% is very respectable for the large majority of players, and means you're well above average yourself.

I couldn't be bothered with picking and choosing opponents! It's first come, first served with me. I presume this ELO system takes account of your opponents (current) rating? The only advantage of the simpler win% is that it can be calculated retrospectively.

Simon
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matt N

Pennsylvania
msg tools
Avatar
mbmbmbmbmb
So I figured out a formula to get the score; it is as follows:

=IF(A2="Stunna : 3.0",3,IF(A2="Stunna : 2.0",2,IF(A2="Stunna : 1.0",1,0)))

Replace Stunna with your name (and A2 with the cell #) and drag the formula across four columns, then sum the values from the four columns to find your score for that game. I get a 73% win rate, although I maintain that win % is a misleading statistic.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Serge Levert
Canada
Vancouver
British Columbia
flag msg tools
badge
Avatar
mbmbmbmbmb
Stunna wrote:
I get a 73% win rate, although I maintain that win % is a misleading statistic.


Win% becomes slightly more meaningful if qualified with average game size. Obviously playing exclusively 2p guarantees a much higher winrate than playing exclusively 6p.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Rob Neuhaus
United States
New York
NY
flag msg tools
Avatar
mbmbmbmbmb
Even average game size is the wrong metric. Consider playing 99 two player games, and 1 10000 player game.

If you played with all equal skills, you'd expect very close to a 50% winning percentage, with an average game size of ~100. It sounds like quite a feat.

 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
David desJardins
United States
Burlingame
California
flag msg tools
badge
Avatar
mbmbmbmbmb
rrenaud wrote:
Even average game size is the wrong metric.


Harmonic mean.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Matt N

Pennsylvania
msg tools
Avatar
mbmbmbmbmb
Er, it's 73% normalized by number of players, so my tendency to get 2nd in 4 player games counts as 2 wins and 1 loss. I still maintain win rate is misleading.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Randall Bart
United States
Winnetka
California
flag msg tools
designer
Baseball been bery bery good to me
badge
This is a picture of a published game designer
Avatar
mbmbmbmbmb
DaviddesJ wrote:
rrenaud wrote:
Even average game size is the wrong metric.


Harmonic mean.

Bob Dylan blows a mean harmonic.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Simon Harris
United Kingdom
Sutton
Surrey
flag msg tools
Avatar
mbmbmbmbmb
... and then I thought (leaving aside the arguments discussions on the relative merits of win% as a meaningful indication of how poor I am at this game!) that a single win when I only have 10 games under my belt would affect the overall % much more than when I'd played 100. So I tried plotting 'running' win%. Choosing an arbitrary look-back period of 10 games (that's matchups, rather than actual games) I stuck both lines on the same graph. Now I can see that my winning-ness definitely changes quite a bit over time even when it *is* evened out over 10 games.


Simon
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.