Batter up! A different perspective of player performance

Actionjick

Forum Guru
Joined
Apr 23, 2020
Messages
7,468
Reaction score
4,992
Location
Kent, Ohio
First name
Darryl
Country
llUnited States
I'm not a fan of player ratings/rankings but like a musical earworm the APCR thread kept occupying my thoughts. Captain Bacchus and I were talking it over and came to the conclusion that rating systems are modeling themselves on the wrong game, chess. I believe it should be baseball, a game that thrives on statistics, and player performance should be modelled on batting average. For those unfamiliar it is a percentage based on the number of a baseball players at bats versus how many hits they made. 1000 at bats, 250 hits and a .250 batting average.

For ASL why not just go by the players won/ loss record somehow taking into account the number of games played?

Example:
Player 7104. 100 games played. 50 wins. A " batting average " of .500

Don. 1000 games played. 650 wins. .650 batting average.

Bobbyjick. 10 games played. 6 wins. .600 batting average

While all 3 players have similar batting averages by looking at the games played you can tell that Don is a consistently good player and much more experienced than the other two. Bobbyjick as a newbie is doing well but hasn't played enough to be able to really rate him as a player. Player 7104 has played a good number of games and we can assume his average is an adequate reflection of his skill.

What I like about this is it doesn't lend itself to the " you're rating is too low for me to play you " syndrome. A win versus a newbie is the same as a win versus one of the Big Guys. In baseball a hit off of pitcher Walter Johnson was recorded the same as a hit off of a rookie pitcher making his first big league start. It's a system that's worked for baseball for over 125 years. The beauty is you can make an accurate comparison of two players perfomances who lived 100 years apart because you are not looking at who they batted against but how well they did when they batted.

In the APCR thread I was questioning what the objective of a rating system was so it's only fair that I do that here for my proposal.

If I were a TD the most important criteria for me would be the number of games someone has played, the player`s experience. Looking at the won/ loss record is helpful but can be misleading, too many unknowns. Looking at our three players even though Bobbyjick has a higher batting average than 7104 I know the obvious matchup is 7104 versus Don. Good luck Don!

I'm sure there are shortcomings to this approach but after all it was proposed by someone who is not a fan of ratings. I like baseball though. Please feel free to rip this idea to shreds. Captain Bacchus and I have thick skins.

Btw, imho should be the Walter Johnson Award, not Cy Young. Apologies to our foreign friends for this diversion into America's national pastime.
 
Last edited:

bendizoid

Official ***** Dickweed
Joined
Sep 11, 2006
Messages
4,630
Reaction score
3,244
Location
Viet Nam
Country
llUnited States
I know of no person whom would refuse to play somebody because of a low rating. I’m sure there are a few but who cares? Why should a fragile ego that places undue value on a ASL rating have a disproportional effect on everybody else? Maybe they don’t like massacring newbies or helping them win and would rather face a challenge. Anybody obsessed with their ASL rating should probably reevaluate their priorities. You’re only as good as the last game you played.
 
Last edited:

Pacman Ghost

Senior Member
Joined
Feb 25, 2017
Messages
590
Reaction score
298
Location
A maze of twisty little passages, all alike
Country
llAustralia
Captain Bacchus and I were talking it over and came to the conclusion that rating systems are modeling themselves on the wrong game, chess.
There is another important distinction between chess and ASL: every single game of chess is (almost) exactly evenly balanced.

If Alice loses against Bob in a game of ASL, it's recorded as a simple win/loss, but what if the scenario is unbalanced 70-30 against Alice? What if we were able to assess how well Alice did, and give her some credit, despite having lost. For example, some scenarios have the concept of a win/decisive win/overwhelming win, or if she managed to hang on until the very last turn, that's much better than losing with 2 turns to spare.

What I like about this is it doesn't lend itself to the " you're rating is too low for me to play you " syndrome. A win versus a newbie is the same as a win versus one of the Big Guys.
If we take scenario balance into account, that could help with this. Give the far stronger player the wrong end of an unbalanced scenario, and he could still get some points, even if he loses.
 

von Marwitz

Forum Guru
Joined
Nov 25, 2010
Messages
14,358
Reaction score
10,207
Location
Kraut Corner
Country
llUkraine
Example:
Player 7104. 100 games played. 50 wins. A " batting average " of .500

Don. 1000 games played. 650 wins. .650 batting average.

Bobbyjick. 10 games played. 6 wins. .600 batting average

While all 3 players have similar batting averages by looking at the games played you can tell that Don is a consistently good player and much more experienced than the other two. Bobbyjick as a newbie is doing well but hasn't played enough to be able to really rate him as a player. Player 7104 has played a good number of games and we can assume his average is an adequate reflection of his skill.
Note, that this model would only (partly) work in a tournament environment.

Imagine two ASL Buddies, Dick and Willie. Both always play each other and never anyone else. Both know hardly a thing about ASL, but yet Dick is better consistently than Willie. They play 100 games. Dick wins 95 and has a 'batting average' of .95

Now Pleva and Fortenberry play 100 games. Pleva manages to win 54 of these, so his 'batting average' is .54 and Fort's .46.

The overall ranking is:

.95 Dick
.54 Pleva
.46 Fortenberry
.05 Willie

Does this represent actual skill?

The model might work if they would be playing in one combined 'league' and not in isolated pockets.
But even if all play in one combined 'league' it does help to arrive at realistic assessments more quickly if the skill level of the opponent is taken into account.

von Marwitz
 
Last edited:

dlazov

Elder Member
Joined
Mar 22, 2004
Messages
7,991
Reaction score
1,377
Location
Toledo, Ohio
First name
Don
Country
llUnited States
This is a fantastic idea! Batting Averages concept makes the mot sense over the AREA or APCR none sense.

Scenario balance is overrated, a dog is a dog anyway.
 

Actionjick

Forum Guru
Joined
Apr 23, 2020
Messages
7,468
Reaction score
4,992
Location
Kent, Ohio
First name
Darryl
Country
llUnited States
Note, that this model would only (partly) work in a tournament environment.

Imagine two ASL Buddies, Dick and Willie. Both always play each other and never anyone else. Both know hardly a thing about ASL, but yet Dick is better consistently than Willie. They play 100 games. Dick wins 95 and has a 'batting average' of .95

Now Pleva and Fortenberry play 100 games. Pleva manages to win 54 of these, so his 'batting average' is .54 and Fort's .46.

The overall ranking is:

.95 Dick
.54 Pleva
.46 Fortenberry
.05 Willie

Does this represent actual skill?

The model might work if they would be playing in one combined 'league' and not in isolated pockets.
But even if all play in one combined 'league' it does help to arrive at realistic assessments more quickly if the skill level of the opponent is taken into account.

von Marwitz
You are quite correct in your assessment as my own batting average would illustrate. I played mainly against Fish, probably over 1000 scenarios. I never did a complete breakdown of all those but did so for 1990 through 1991. I can't recall the number of scenarios played during that period but I remember that my winning percentage was slightly over 25% so roughly a .250 batting average. Fish was an exceptional gamer so I'm quite satisfied with that average although I would be the first to admit too many of those victories could be attributed to hot dice and not skill. I'm not sure of my average against other opponents but it was certainly higher than .250. Still by looking at number of games played and winning average a TD could conclude I was a relatively skilled player and match me accordingly.

Does AREA or it's cousins accurately reflect skill? I'm not sure. This approach is based more on experience. Willie and Dick have played 100 scenarios so must have learned something about the game. Fort and Pleva have certainly played more than 100 games and that would be the determining factor of how I would arrange matchups.

To look at the previous examples I would arrange the matchups this way:

Pleva vs. Don
Dick vs. 7104
Bobbyjick vs. Willie

Perhaps not ideal but you work with what you have.

Once again major kudos to you for your Death Star thread. Keep them coming!

Btw I was way off on my scenarios vs. Fish. Probably over 2500.
 
Last edited:

Actionjick

Forum Guru
Joined
Apr 23, 2020
Messages
7,468
Reaction score
4,992
Location
Kent, Ohio
First name
Darryl
Country
llUnited States
I know of no person who would refuse to play somebody because of a low rating. I’m sure there are a few but who cares? Why should a fragile ego that places undue value on a ASL rating have a disproportional effect on everybody else? Maybe they don’t like massacring newbies or helping them win and would rather face a challenge. Anybody obsessed with their ASL rating should probably reevaluate their priorities. Your only as good as the last game you played.
The correct attitude imho. Still, from previous posts this is not a universal attitude, unfortunately.
 

Actionjick

Forum Guru
Joined
Apr 23, 2020
Messages
7,468
Reaction score
4,992
Location
Kent, Ohio
First name
Darryl
Country
llUnited States
This is a fantastic idea! Batting Averages concept makes the mot sense over the AREA or APCR none sense.

Scenario balance is overrated, a dog is a dog anyway.
Captain Bacchus and I thank you for your kind words.

Personally I think the concept of a balanced scenario is silly and actually contrary to what ASL is trying to simulate. War is not balanced. Besides it's more fun playing as the underdog and a victory in such circumstances was more satisfying to me.
 

Sparafucil3

Forum Guru
Joined
Oct 7, 2004
Messages
11,335
Reaction score
5,071
Location
USA
First name
Jim
Country
llUnited States
There is another important distinction between chess and ASL: every single game of chess is (almost) exactly evenly balanced.
Not only is it nearly balanced, the imbalance which exists is well know. Additionally, every time my pawn takes your pawn, the outcome isn't decided by a random DR. Chess simply is not a good basis for comparison to ASL. -- jim
 

dlazov

Elder Member
Joined
Mar 22, 2004
Messages
7,991
Reaction score
1,377
Location
Toledo, Ohio
First name
Don
Country
llUnited States
The only main problem with any rating system and the fact that they are all flawed is that any rating system needed to be implemented from day one of the game to day one of the first time you played the game.


I am sure I have had hundreds of games (if not more) since 1985, also during that three year stretch that Mike and I played each other exclusively in RB CGs generating 19-30 scenarios per CG played, how would that be included?

Plus what about all the play tests I did for LC and BFP and with Brian M (Tooz) and his creations?
 

Actionjick

Forum Guru
Joined
Apr 23, 2020
Messages
7,468
Reaction score
4,992
Location
Kent, Ohio
First name
Darryl
Country
llUnited States
Batting Average?
Batting Average?!?

Are you stuck in the 20th Century.

Park Adjusted OPS is a much better performance metric.
Lol, I have no idea what that is.
Considering the current state of affairs the 20th century isn't looking too bad. Strike that thought, it was a pretty awful century also.
 

Actionjick

Forum Guru
Joined
Apr 23, 2020
Messages
7,468
Reaction score
4,992
Location
Kent, Ohio
First name
Darryl
Country
llUnited States
Batting Average?
Batting Average?!?

Are you stuck in the 20th Century.

Park Adjusted OPS is a much better performance metric.
Checked it out on the Google, very interesting. Thanks for the new perspective!
 

Robin Reeve

The Swiss Moron
Staff member
Moderator
Joined
Jul 26, 2003
Messages
19,595
Reaction score
5,557
Location
St-Légier
First name
Robin
Country
llSwitzerland
Personally I think the concept of a balanced scenario is silly and actually contrary to what ASL is trying to simulate. War is not balanced. Besides it's more fun playing as the underdog and a victory in such circumstances was more satisfying to me.
I would dare suggest that there is a sensible difference between an unbalanced situation ("war isa not balanced") and a scenario that doesn't offer victory conditions which aim to be challenging for both sides.
I think that a designer - and the team of playtesters - should do his best to aim that type of balance.
For an example, if the "underdog" normally should be wiped out in six turns, then a good design would require that he holds on seven turns.

Otherwise, scenario design would be summed up by throwing a bunch of counters haphazardly on a random set of boards with nonsenical VCs.
What some TPP indeed do, in some cases in a systematical way.

As much as balance is a fuzzy, elusive notion, spending a number of hours just to see a desperate situation unfold with no real tactical challenge is, for me, a real waste of time.
I love to play underdogs and to have to solve the problematic of asymetrical combat.
But heaping twelve Tiger tanks and twenty SS 658 squads in fortified buildings led by three 10-3 leaders vs. two Russian conscript HS in open ground and defining VC as having the Russian destroy 10 of the Tigers and capture or eliminate all the 658 is not what I would call fun nor a good design.
I would not like to play the German nor the Russian in that case.
 

Actionjick

Forum Guru
Joined
Apr 23, 2020
Messages
7,468
Reaction score
4,992
Location
Kent, Ohio
First name
Darryl
Country
llUnited States
I would dare suggest that there is a sensible difference between an unbalanced situation ("war isa not balanced") and a scenario that doesn't offer victory conditions which aim to be challenging for both sides.
I think that a designer - and the team of playtesters - should do his best to aim that type of balance.
For an example, if the "underdog" normally should be wiped out in six turns, then a good design would require that he holds on seven turns.

Otherwise, scenario design would be summed up by throwing a bunch of counters haphazardly on a random set of boards with nonsenical VCs.
What some TPP indeed do, in some cases in a systematical way.

As much as balance is a fuzzy, elusive notion, spending a number of hours just to see a desperate situation unfold with no real tactical challenge is, for me, a real waste of time.
I love to play underdogs and to have to solve the problematic of asymetrical combat.
But heaping twelve Tiger tanks and twenty SS 658 squads in fortified buildings led by three 10-3 leaders vs. two Russian conscript HS in open ground and defining VC as having the Russian destroy 10 of the Tigers and capture or eliminate all the 658 is not what I would call fun nor a good design.
I would not like to play the German nor the Russian in that case.
Yes you are correct but as most things in ASL this is conditional. I would play a 70/30 or 80/20 scenario as the underdog in a friendly situation as long as the scenario was interesting or fun. At a tournament I would prefer it to be closer to 60/40 or 50/50.

In your example just think of the great story the improbably victorious Russian could tell. Even a close defeat would would be glorious.

I seem to recall a scenario where a handful of vehicle counters were dropped on the board as wrecks. Anyone have a recollection of this?
 

Eagle4ty

Forum Guru
Joined
Nov 7, 2007
Messages
6,913
Reaction score
5,094
Location
Eau Claire, Wi
Country
llUnited States
...

I seem to recall a scenario where a handful of vehicle counters were dropped on the board as wrecks. Anyone have a recollection of this?
IIRC it was Something like Ride of The Valkyries from Jeff Shields (JDS scenarios) or at least one of that bunch about the Cherkassy Pocket action.

Do you remember the one where the guy said to cut off the unused portion of the board with an Exacto knife because you didn't need the entire board when playing the scenario?o_O:LOL:
 

Perry

Forum Guru
Joined
Feb 4, 2003
Messages
1,810
Reaction score
2,762
Location
Baltimore, MD
Country
llUnited States
I seem to recall a scenario where a handful of vehicle counters were dropped on the board as wrecks. Anyone have a recollection of this?
58 Ci Arrendiamo from Hollow Legions, where eight German AFV are simultaneously dropped from one foot above a specific hex and then positioned as wrecks in their landing hexes.

Some circles were advocating for having the Balance to be to drop them from six inches.
But it wasn't clear who should do the dropping. ;)
 

Actionjick

Forum Guru
Joined
Apr 23, 2020
Messages
7,468
Reaction score
4,992
Location
Kent, Ohio
First name
Darryl
Country
llUnited States
58 Ci Arrendiamo from Hollow Legions, where eight German AFV are simultaneously dropped from one foot above a specific hex and then positioned as wrecks in their landing hexes.

Some circles were advocating for having the Balance to be to drop them from six inches.
But it wasn't clear who should do the dropping. ;)
Ah, good job. One of the few Italian phrases I know. Loved the HL box art!
 

Actionjick

Forum Guru
Joined
Apr 23, 2020
Messages
7,468
Reaction score
4,992
Location
Kent, Ohio
First name
Darryl
Country
llUnited States
Not only is it nearly balanced, the imbalance which exists is well know. Additionally, every time my pawn takes your pawn, the outcome isn't decided by a random DR. Chess simply is not a good basis for comparison to ASL. -- jim
Now Captain Bacchus and I are trying to figure out how to incorporate dice rolling into chess. This forum can lead to too much thinking and not enough drinking.
 
Top