G40 League House Rule project

[DarthMaximus]

This thread is for Gamerman01 to use to update his rankings.

Gamer, just post your criteria in your first post and you can start/update whenever you want.

–--------------

Please note, that these rankings have no barring on the league and will not be taken into account for league playoffs.

[gamerman01]

Thanks a lot, Darth.
That’s right.  This thread is just for fun, and has no bearing on League regular season or playoff play.
These rankings are unofficial, and subjective.

1)  Regular season standings with W/L records will be posted regularly here.  These should always match JWW’s, but I plan to post much more frequently.
2)  I will regularly upload my Excel spreadsheet of the matrix of results, so players can see who has played who, and who won the matchups, at a glance.
3)  I will post my subjective point rankings, which are designed to reflect strength of opponents.  (Because players can choose to play who-ever, and there is no schedule)

Here is the methodology for the point rankings:

1)  I will consider the 2012, 2011 (and possibly 2010) league standings in determining quality of wins/losses from 2012 league results.
2)  I will divide all league participants into 3 tiers, based on past records (2012, 2011, and possibly 2010)
3)  I will award points for wins.  3 points against top tier, 2 points against mid-tier, and 1 point against lower tier.
    I will deduct points for losses.  1 point against top tier, 2 points against mid-tier, and 3 points against lower tier.

Note that your point rankings will change not only when you complete a game, but they could change up or down if you or your past opponents move up or down tiers.

I welcome comments on the standings as they are posted throughout the year.  I will be posting the point rankings and spreadsheet for the 2011 season when it is finalized here in the next couple of days, for your enjoyment and/or curiosity.  For the sake of consistency, I do not plan to change the point ranking system during the year, but will consider suggestions or comments for improvement for subsequent years.

To reiterate, this ranking thread is for fun, and nothing else.  Enjoy.

[gamerman01]

@Gamerman01:

2)  I will divide all league participants into 3 tiers, based on past records (2012, 2011, and possibly 2010)

Tiers will cutoff at 66% and 33% win pct, but there may be exceptions.  If a participant’s past point ranking is significantly different than his or her straight win percentage, they may be assigned to the different tier.  I will note any such re-assignments with the ranking posts.

but they could change up or down if you or your past opponents move up or down tiers.

Check that - only if your past opponent(s) move up or down.  If you move up or down a tier, it would not affect your own point ranking (too bad I can’t edit on this thread)

[DarthMaximus]

I’d suggest ignoring 2010 due to bid levels. 
I’m not sure those results would hold much weight now.  Infact for our first season here we even bidded for the Axis.

[gamerman01]

@DarthMaximus:

I’d suggest ignoring 2010 due to bid levels. 
I’m not sure those results would hold much weight now.  Infact for our first season here we even bidded for the Axis.

OK, thanks!  Was thinking I might take into consideration tournament results along with 2011 league…
Anyway, if anyone is in a different tier than the 2011/2012 combined record indicates, I will note that along with the reason.
Thanks again.

[DarthMaximus]

2011 should be good, but taking a quick look back at 2010, it has most of the same names (yourself, Yoshi, DD, Zuck, etc.), but bid levels were probably much lower.

Not sure how much “extra” work this would be, but you could do two columns

Current rankings (takes 2011 and then 2012 results as they come in) and then

3 year rankings or “life tme” rankings (takes 2010, 2011 and then 2012 results)

But you’ll probalby end up with the same names at the top.  I don’t think anyone really dropped off from 10 to 11.  Also some of those members may no longer be that active (Hobo for one)

[gamerman01]

Yes, I also think 2010 has some valuable information regardless of bid levels (if players took both sides, it would even out anyway).  More information is normally better than less….

As I post stuff, I’ll describe my thinking and my methods as appropriate!

[Guest]

Sounds like a lot of time and effort. Thanks for doing it!

[gamerman01]

@Ol’:

Sounds like a lot of time and effort. Thanks for doing it!

No, it really isn’t much!  You’re welcome!

[gamerman01]

2011 League standings have been finalized by JWW, so I will be posting the final spreadsheet and point rankings soon.

[gamerman01]

Final spreadsheet/matrix attached for 2011 league

I reviewed 2010 standings but made no changes (mostly different players anyway)
Two exceptions to 66%, 33% tier cutoffs for these standings:
  BoldFresh adjusted to Tier 1 because of his point total (strong “schedule”)
  Darth Maximus (at 33%) adjusted to Tier 2 for same reason (All wins against Tier 1 and Tier 2 players)

The effect of these 2 adjustments is to add 1 point to each player who played Bold or Darth for each game played.
Standings are sorted by point total on the Excel spreadsheet (2003 version so that it may be uploaded).

Significant observations:
#1, #2, #3 and #4 are exactly the same as JWW’s pct. standings
Tyzoq drops to #7.  Note only 1 game played vs. Tier 1 player

Boldfresh jumps from #8 to #5.  This is Exhibit A for the merit of the strength of schedule point ranking system.  Look at the details of his results.  10 games played against Tier 1 players.  11 games vs. Tier 2 players.  Only 5 against Tier 3 players.
(Actually he would be at #4 with 13 points if I make Botider Tier 2.  Just noticed Botider’s 3 games were against Bold, a Tier 1 player)

Dragon and Billy have the same point totals, while Dragon is 7-3 and Billy is 6-5.  Billy played 5 games vs. Tier 1 while Dragon only played 2 against top players.

For 2011 there were few significant differences between pct. standings and SOS standings because most players in 2011 played a good balance of top, middle, and lower tiered opponents.

Point rankings 11 3 11.xls

[Tyzoq]

@Gamerman01:

Significant observations:
#1, #2, #3 and #4 are exactly the same as JWW’s pct. standings
Tyzoq drops to #7.  Note only 1 game played vs. Tier 1 player

Boldfresh jumps from #8 to #5.  This is Exhibit A for the merit of the strength of schedule point ranking system.  Look at the details of his results.  10 games played against Tier 1 players.  11 games vs. Tier 2 players.  Only 5 against Tier 3 players.
(Actually he would be at #4 with 13 points if I make Botider Tier 2.  Just noticed Botider’s 3 games were against Bold, a Tier 1 player)

What I’m beating him down in our game, with 966 TUV difference between us!  :P

Seriously though, I do feel out of place on JWWs standings sitting between you and Zhukov.  Nice work I look forward to seeing the final results

[dutchmand]

@Tyzoq:

Seriously though, I do feel out of place on JWWs standings sitting between you and Zhukov.  Nice work I look forward to seeing the final results

That’s nonesense. You are one hell of a player Tyzok.  You can hang with anyone in the league.  :-P

[boldfresh]

@Tyzoq:

@Gamerman01:

Significant observations:
#1, #2, #3 and #4 are exactly the same as JWW’s pct. standings
Tyzoq drops to #7.  Note only 1 game played vs. Tier 1 player

Boldfresh jumps from #8 to #5.  This is Exhibit A for the merit of the strength of schedule point ranking system.  Look at the details of his results.  10 games played against Tier 1 players.  11 games vs. Tier 2 players.  Only 5 against Tier 3 players.
(Actually he would be at #4 with 13 points if I make Botider Tier 2.  Just noticed Botider’s 3 games were against Bold, a Tier 1 player)

What I’m beating him down in our game, with 966 TUV difference between us!  :P

Seriously though, I do feel out of place on JWWs standings sitting between you and Zhukov.  Nice work I look forward to seeing the final results

Tyzoq IS way ahead in TUV at the expense of having almost no tech.  so my TUV is really a lot closer, and also the total TUV in the game right now is 3,869 Axis - 2,973 Allies.  I don’t know how much the tech actually pumps up my TUV but it’s probably considerable, say 300-400 IPC.  then you are only looking at a spread of 15-20% maybe.  allies have 22 techs (3 are russian, whos capital is in Axis hands) and the Axis has 4 techs.

this game was started on December 16th of 2010 and is nearing 100 pages and entering Round 49.  should be interesting to see how this one plays out, we have agreed it will not be a league result, now we are just playing for the glory!  i don’t know what the longest A&A game ever played has been but this one could be getting in the neighborhood  :-o

[gamerman01]

@Tyzoq:

Nice work I look forward to seeing the final results

Thank you -
You did see the spreadsheet to download, right?  Those are the “final results” for 2011.

I am glad for the attention my point rankings are getting - thanks all, for your posts.  It will be interesting during 2012 to have constantly updated standings and adjusted (point) rankings, I think.

[Tyzoq]

@Gamerman01:

Thank you -
You did see the spreadsheet to download, right?  Those are the “final results” for 2011.

I did not see that, thanks for pointing it out.

[gamerman01]

@Boldfresh:

Tyzoq IS way ahead in TUV at the expense of having almost no tech.  so my TUV is really a lot closer, and also the total TUV in the game right now is 3,869 Axis - 2,973 Allies.  I don’t know how much the tech actually pumps up my TUV but it’s probably considerable, say 300-400 IPC.  then you are only looking at a spread of 15-20% maybe.  allies have 22 techs (3 are russian, whos capital is in Axis hands) and the Axis has 4 techs.

Always helps to hear both sides of the story!
19 to 4 tech disparity - that definitely makes up for a chunk of the TUV disparity.

I stand behind my rankings that result in 8 points for Tyzoq.  Of course Dutchman’s going to talk him up, Dutch you’re the only “statement win” that Tyzoq had last year!  :-) :-)
At 5-1 with only 1 win against a very successful player, you shouldn’t be ranked extremely high - so I stand by my system.  :-)  In other words, I believe a player could be much better than his ranking on the point system indicates.  You have to prove it, for the points to come!

Don’t get me wrong - I appreciated your post, DD, and hope you continue to comment on this thread from time to time.

Our league is a bit like college football.  Everybody doesn’t play everybody else on a schedule like in Professional sports.  Also, a “season” may only be 8-12 games.  Most players could beat most other players given the right circumstances, so there are “upsets”.  The point ranking system isn’t perfect (you should always be able to find over or under-rated players, as in College FB), but I’ve already seen it gives us insights that we wouldn’t have without it.

[Tyzoq]

@Gamerman01:

At 5-1 with only 1 win against a very successful player, you shouldn’t be ranked extremely high - so I stand by my system.

I just wanted to make sure it’s clear that I appreciate the work you’ve done.  And I do not have any problems with your system.  I apologize if I came across as though I did.  Like I said I do not feel in anyway that I am as good as Zhukov, Dutchman, Lucky Lindy, Rising Dragon, etc.

PS.  Thanks for the comment Dutchman.

[gamerman01]

@Tyzoq:

I just wanted to make sure it’s clear that I appreciate the work you’ve done.  And I do not have any problems with your system.  I apologize if I came across as though I did.  Like I said I do not feel in anyway that I am as good as Zhukov, Dutchman, Lucky Lindy, Rising Dragon, etc.

Oh, no, I didn’t take it that way - thanks for making sure.

[DarthMaximus]

Thanks for all the info Gamerman.  Good stuff!

@Gamerman01:

Boldfresh jumps from #8 to #5.  This is Exhibit A for the merit of the strength of schedule point ranking system.  Look at the details of his results.  10 games played against Tier 1 players.  11 games vs. Tier 2 players.  Only 5 against Tier 3 players.

I think another reason Bold jumped up, is just in the number of games played.  With the exception of OBG and Soul, Bold doubled the amount of games played by any other player.  I think that is another reason why Tyzoq dropped, its not necessarily due to the quality of opponent as it is he was only able to finish 6 games.  Infact Tyzoq is the highest rated player of anyone with less than 10 games.

Assuming you maintain a winning percentage of over 50% and only play tier 2 players.  It will always be to your advantage to play more and more games.  You gain 2 for a win but lose 2 for a loss, but as long as you win over 50%, it pays to play more and more games.  If you play 20 games and go 11-9, you’ll end up with 4 pts, if you play 40 and have the same win% (22-18) you end up with 8 pts.  Both players have a 55% but I’m not sure player 2 is better.
I suppose you could argue it is harder to win 22 games (out of 40) against tier 2 players than it is to win just 11 (out of 20) from tier 2.

I’m wondering if you considered dividing pts by games played to come with with a players average pts per game?

I think a lot of this has to do with some relatively small samples and I think you even mentioned you had trouble rating players with only a few games.

Also note, I’m not criticising the ranking of Bold, I just used him b/c he played the most games.  Bold is infact a really good player and worthy of a jump in the rankings.

Not to get too stats-y, but to steal from Baseball and something like wins above replacement, I think the fact that OBG went 11-11 and yet still had positve pts shows he’s played slightly higher than avg players.

For example if you use tier 2 (since its the avg of tier 1 and 3) as the base line you can look at your numbers:
Gamerman 12-1 and assume you had only played tier 2 you’d be at 22 pts.  But your total is 21.  1 below replacement (or -1) Meaning you may not have played the strongest schedule.
Dutchman 8-2 would be at 12 pts (tier 2), but ended up with 14.  +2 or 2 above replacement.
Billy 6-5 would be 2 pts but ended with 7.  +5 above replacement, signifying he played a strong schedule and did well.
Tyzoq 5-1 = 8 pts but and 8.  Meaning he played an average schedule.
Bold 15-11 = 8 pts but had 10.  Confirms he played a tough schedule.

Intersting stuff.  I think I could go on all day.  lol!

Please note, I’m not critising your rankings, but I’m numbers junkie as well, just looking at different ways to use your numbers.  8-)