Home Field Advantage?

[Scott Gustafson]

god knows my team needs it... we are 21-34 at home.

[Big Willis 16]

Can someone chime me in on "NDA?"

[fasteddie_21]

Originally posted by Big Willis 16
Can someone chime me in on "NDA?"

Me thinks that it stands for "Non-Disclosure Agreement"...I think.

-fe

[MJohnson]

Here it is, a two team league where post teams are none other than the 1994 San Diego Padres.

I set up a 200 game series with trades, injuries, and fatigue off. All games were played with "Home team" (clever, I know)in the field first.

The home team went 106-94 for a .530 winning pct. That is reasonably close to the 55% mentioned by Eugene Church above.

The interesting point is that the Home team had a 95-105 Pyth record. I think this is due to the fact that the Home team batted in about 100 less innings since they were home every time.

<pre><font size="+1">
Team W L PCT GB AVG HR R AB H 2B 3B BB K OBP SLG OPS SB
Home Team 106 94 .530 - .276 147 906 7036 1942 366 32 504 1188 .325 .400 .725 132
Away team 94 106 .470 12.0 .270 167 954 7320 1980 376 43 560 1133 .323 .402 .726 123


Team W L PCT GB ERA S IP HA R ER HRA BB K OAVG CG SHO
Home Team 106 94 .530 - 4.24 57 1861.0 1980 954 877 167 560 1133 .270 6 1
Away team 94 106 .470 12.0 4.07 50 1765.0 1942 906 798 147 504 1188 .276 12 4
<font size="-1">
</pre>

[Dargone]

I too would like to know how this is implemented...

[Henry]

The method used is probably is not going to be announced - certainly not by the Beta Team because of the NDA.

Henry

[BasicGuy]

Henry are you going to have It turned on? if so Its good enough for me.

[Henry]

Originally posted by BasicGuy
Henry are you going to have It turned on? if so Its good enough for me.

You bet

Henry

[gmo]

I was quite excited to see home field advantage be implemented. Yet another small but fantastic example of how great OOTP is, responding to its customer base. Anyway, I have played around with HFA for a while now, including with OOTP6, and now V6.10.

But I have a pair of related questions - Is it working correctly if enabled in V6.10?/Was it automatically enabled in V6.00?

Let me explain why I ask. The long-term historical average for home teams is ~.540, so that may have been the goal of the option but maybe not. Initially with OOTP6 I made a 6 team replay league where the teams were identical, the rosters the same and the players editted to have all the same ratings. Three teams play only home games, three play only away games. I ran that league for 50 seasons expecting, since I assumed there was no home field advantage in V6.00, for home teams to win half the games. Home teams actually won at a .511 clip. The home team stats were slightly better (e.g., +.007 in OPS and -0.07 in ERA), but I figured it was just a quirky sample giving me on average slightly better home teams. Then I tweaked some player ratings to make the home teams better-rated in the game (since MLB data shows home teams have better overall stats in some categories) and ran another 50 seasons. Home teams won at a .522 rate in that experiment, with the home/away stat differences not quite up to that seen in 2001-2003 MLB data. At that point I'm thinking that at some point in the future, as had been suggested before, this could be a good route to implementing HFA.

Then today... Immediately after grabbing V6.10 I reran the initial setup with all 6 teams being identical, and the 50 season result had the home teams winning pct at .510, essentially the same as in the initial version of the game. And the average statistical differences between the home and away teams was virtually the same as the first case. I double-checked, and I did have the HFA option enabled on the engine setup screen for the league. Each of the 3 experiments had almost 25000 team games, so the sample is not tiny even if it is still not very big.

I'm sure others are also looking at this new option now, so I am eager to hear more results.

Oh, and since I haven't done it anywhere else yet let me also echo the common sentiment generally piled in the release threads to Markus and OOTP - Great job, your work and dedication is much appreciated.

[KurtBevacqua]

I think the best way to implement a home field advantage would be having guys tire out quicker on the road than at home. The most advantage home teams should have, aside from knowing the park better, is they aren't living out of a suitcase and should be less fatigued. Now if someone can only figure out what the heck is happening to guys who are tired.

[potatoman]

Originally posted by gmo
I was quite excited to see home field advantage be implemented. Yet another small but fantastic example of how great OOTP is, responding to its customer base. Anyway, I have played around with HFA for a while now, including with OOTP6, and now V6.10.

But I have a pair of related questions - Is it working correctly if enabled in V6.10?/Was it automatically enabled in V6.00?

Let me explain why I ask. The long-term historical average for home teams is ~.540, so that may have been the goal of the option but maybe not. Initially with OOTP6 I made a 6 team league where the teams were identical, the rosters the same and the players editted to have all the same ratings. Three teams play only home games, three play only away games. I ran that league for 50 seasons expecting, since I assumed there was no home field advantage in V6.00, for home teams to win half the games. Home teams actually won at a .511 clip. The home team stats were slightly better (e.g., +.007 in OPS and -0.07 in ERA), but I figured it was just a quirky sample giving me on average slightly better home teams. Then I tweaked some player ratings to make the home teams better-rated in the game (since MLB data shows home teams have better overall stats in some categories) and ran another 50 seasons. Home teams won at a .522 rate in that experiment, with the home/away stat differences not quite up to that seen in 2001-2003 MLB data. At that point I'm thinking that at some point in the future, as had been suggested before, this could be a good route to implementing HFA.

Then today... Immediately after grabbing V6.10 I reran the initial setup with all 6 teams being identical, and the 50 season result had the home teams winning pct at .510, essentially the same as in the initial version of the game. And the average statistical differences between the home and away teams was virtually the same as the first case. I double-checked, and I did have the HFA option enabled on the engine setup screen for the league. Each of the 3 experiments had almost 25000 team games, so the sample is not tiny even if it is still not very big.

I'm sure others are also looking at this new option now, so I am eager to hear more results.

Oh, and since I haven't done it anywhere else yet let me also echo the common sentiment generally piled in the release threads to Markus and OOTP - Great job, your work and dedication is much appreciated.

Did you have injuries, trades, fatigue off?

[gmo]

Originally posted by potatoman
Did you have injuries, trades, fatigue off?

Sorry, should have specified that. Yes, everything was off - minors, finances, trades, injuries, fatigue... Am I missing anything? Only players in the league were on the 25 man major league rosters, and I checked in the very beginning that the rotations and lineups were all identical and that when all were FA before being assigned, the 6 versions of each player were clustered together in the Available FA HTML report.

[potatoman]

Originally posted by gmo
Sorry, should have specified that. Yes, everything was off - minors, finances, trades, injuries, fatigue... Am I missing anything? Only players in the league were on the 25 man major league rosters, and I checked in the very beginning that the rotations and lineups were all identical and that when all were FA before being assigned, the 6 versions of each player were clustered together in the Available FA HTML report.

I'n this case, I do not thing a 50 year sample would work. Players on each team, even though starting with the same ratings will develop differently causing some players to drop in rating or increase. This would produce different teams basically and weaker teams will lose at home reguardless of HFA. Also, after 10 seasons, most of the teams you have will have completely different players. There will be a new set of young players and none of them would be equal then. Unless I misread what you did and you did 1 season, 50 times....

[gmo]

Originally posted by potatoman
Unless I misread what you did and you did 1 season, 50 times....

Don't think you misread, I just did not say again. It was a replay league. Let me fix that up there...

[potatoman]

Originally posted by gmo
Don't think you misread, I just did not say again. It was a replay league. Let me fix that up there...

So basically it just played out 50 seasons in a row of the same players, same ratings....etc?

[gmo]

Originally posted by potatoman
So basically it just played out 50 seasons in a row of the same players, same ratings....etc?

Exactly. League just resets to starting date at the end of each season (after I push the "reset league" button).

[Le Grande Orange]

You need to run A LOT of seasons to get the overall values correct. Even different 10 year spans can have quite different values. Consider the following:

1901-1910 = NL .5341, AL .5694
1911-1920 = NL .5365, AL .5375
1921-1930 = NL .5481, AL .5454
1931-1940 = NL .5485, AL .5488
1941-1950 = NL .5404, AL .5512
1951-1960 = NL .5437, AL .5348

Aggregate total for 1901-1960 = NL .5420, AL .5478

Now looking at it in terms of teams doing better at home, worse at home, or the same, you get this:

1901-1910 = NL 67-11-2, AL 75-5-0
1911-1920 = NL 65-14-1, AL 66-14-0
1921-1930 = NL 68-11-1, AL 72-8-0
1931-1940 = NL 68-12-0, AL 71-8-1
1941-1950 = NL 66-13-1, AL 67-11-2
1951-1960 = NL 68-10-2, AL 61-16-3

Aggregate total for 1901-1960 = NL 402-71-7, AL 412-62-6.

As one would expect, there is a lot of variability even over longer periods, let alone season to season. However, over the very long term, it works out to around .540 at home average and 84% of teams that will end the season with a better home winning percentage than away winning percentage (that's five out of six teams).

It is interesting to see how it varied based on position of finish as well. Below is a summary for 1901-1960, which is a nice period to study since for those 60 years the league sizes were stable at 8 teams each for the entire time.

        National League 1901-1960
Finish Overall Home   Away   Diff.  Win %
-----------------------------------------
 1st   .6367  .6672  .6059  .0613   4.79
 2nd   .5902  .6277  .5526  .0751   6.35
 3rd   .5603  .6001  .5200  .0801   7.11
 4th   .5229  .5702  .4754  .0949   9.05
 6th   .4426  .4855  .3997  .0858   9.70
 7th   .4062  .4513  .3623  .0891  11.11
 8th   .3541  .3975  .3108  .0867  12.26
-----------------------------------------
TOTAL  .5000  .5420  .4580  .0839   8.39

        American League 1901-1960
Finish Overall Home   Away   Diff.  Win %
-----------------------------------------
 1st   .6397  .6893  .5895  .0998   7.75
 2nd   .5915  .6399  .5430  .0969   8.18
 3rd   .5574  .6077  .5067  .1010   9.03
 4th   .5236  .5660  .4809  .0852   8.10
 5th   .4861  .5325  .4397  .0927   9.54
 6th   .4501  .5004  .4002  .1002  11.17
 7th   .4061  .4582  .3542  .1040  12.85
 8th   .3410  .3829  .2997  .0832  12.29
-----------------------------------------
TOTAL  .5000  .5478  .4522  .0955   9.55

It's interesting to see that in the NL the first place finishers tended to have a much lower differential when compared to other positions, while in the AL it was much more consistent throughout the positions of finish.

[kq76]

LGO, I don't quite understand what is meant by Win % in your table. I noticed the totals for it and Diff. are the same though. I'm sure I'll feel like an idiot after I'm told.

[Le Grande Orange]

Whoops, double post. See post below for explanation.

[Le Grande Orange]

It's a measure of how many more wins, on a percentage basis, a team is getting at home than it would be expected to have based on its overall winning percentage.

Consider a team which finishes 98-56 .636 overall. At home it goes 52-25 .675 and 46-31 .597 on the road. If you take the overall winning percentage and applied it to just the home games, that would yield 49 wins (.636 x 77). The team actually won 52 games at home, however. So, 52 / 49 = 1.0612, which is 6.12%. Therefore, the team won 6.12% more games at home than its overall winning percentage suggested it would.

Hope that makes it a bit clearer.