The predecessor post to this one Prep Times & Game Outcomes: An Analysis By The Numbers analyzed game outcomes, via a data warehouse, in terms of the amount of time teams had to prepare for the game. There were some further questions raised by some of those who commented on the post which I thought were worthy of answering in a post of their own.The findings in the first post were:
- Having an extra 7 days of prep time, which usually occurs when a team is coming off of
it's bye week is worth about 3 points the same as home field advantage. I dubbed this
the "bye prep" advantage.
- These extra days of prep time are not handed out by the schedule makers to all teams equally,
certainly not in any given year, and not even over the last 10 years.
BGN'er and statistics professor Aggie Mo weighed on on the bye prep results in the comments section of the post:
I am a statistics professor, and looking at the 7 day difference, the team with the longer break won 153 and lost 115. If there was no effect from the extra time off, we would expect the winning percentage to be 50-50 since every team has a bye. Therefore, we can calculate a p-value for testing this hypothesis using the binomial distribution, which computes what is the probability of getting a 153-115 W/L split or something more extreme if the assumption that there is no bye effect is true.
The p-value comes out to 0.009. This means that if there were no bye effect, then the probability of getting data like this is < 1%. Traditionally, any p-value less than 0.05 is considered to be “statistically significant”, so in this case our result is considered “statistically significant”, and we can conclude that a 7-day layoff gives an advantage.
If we combine the data from 6, 7, and 8 days together, then this split comes to 181 wins and 138 losses, which corresponds to a p-value of 0.007. Thus, you can make the conclusion that a layoff difference of ~1 week does confer a slight advantage to the team with the extra rest.
Please read the comments section of the earlier article for the entire quote in context.
A number of those who commented (KByars, Talon Talent, DerbyGuy) wanted to see the "Outcomes By Discrepancy In Prep Time Summary" data broken down by Home/Away games, so here it is! Remember, you can click on the numbers in the table to see the detail on the games the numbers represent.
|Outcomes By Discrepancy And Location|
|Discrepancy||Advantaged Overall Wins||Advantaged Home Wins||Advantaged Away Wins||Disadvantaged Overall Wins||Disadvantaged Home Wins||Disadvantaged Away Wins|
7 Day Discrepancy Prep Time Games Overall Results By Location
This section discusses the numbers found in the 7 day discrepancy in prep time row. This is the "bye prep" discrepancy that was shown to be significant in the earlier article.
Advantaged Home Wins - 100/268 = 37.31% Disadvantaged Home Wins - 59/268 = 22.01% Disadvantaged Away wins - 56/268 = 20.89% Advantaged Away Wins - 53/268 = 19.77%
Overall, advantaged home teams won almost twice as many games as any other category. This surprised me, I didn't expect the gap to be that large.
Games Where Advantaged Teams Won
Let's first look at the set of games where the advantaged teams won the games. Overall, if the usual Home/Away (57.374%/42.626%) split was applied to these 153 games the home team would be expected to win 87.73 games and the away team would be expected to win 65.27 games
Advantaged Home Wins As A Percentage Of Advantaged Overall Wins 100/153 = 65.35%
If your team has the bye prep advantage, you want them to play at home. 65.35% of teams in this scenario won their games, as compared to only 57.374% of games won overall by the home team. The bye prep advantage seems to work in conjunction with the home field advantage making it even more likely the advantaged team will win the game.
Advantaged Away Wins As A Percentage Of Advantaged Overall Wins 53/153 = 34.64%
On the other hand, having the bye prep advantage doesn't seem to confer extra wins if your team is playing a road game. The 34.64% win percentage in this context compares unfavorably to to the 42.626% of games won by road teams overall. This result runs contrary to what my intial expectations were.
Home teams did better then I expected in this context, away teams did worse.
Games Where Disadvantaged Teams Won
Next, we'll look at the set of games where the disadvantaged teams won the games. Overall, if the usual Home/Away split was applied to these 115 games the home team would be expected to win 65.941 games and the away team would be expected to win 49.059 games.
Disadvantaged Home Wins As A Percentage Of Disadvantaged Overall Wins 59/115 = 51.30%
If your team has the disadvantage in bye prep time, you still want the game to be at home. The bye prep disadvantage makes it 6.074 percent less likely for them to win the game when compared with the usual home field advantage of 57.374%
Disadvantaged Away Wins As A Percentage Of Disadvantaged Overall Wins 56/115 = 48.69%
On the other hand, if your team has a road game and a bye prep disadvantage as well, your team has a 6.064% better chance of winning the game when compared to their chances of winning a road game overall. (48.69% - 42.626%) This is another result that surprised me.
I would expect that the home field advantage would erode if the home team also had the bye prep disadvantage, and the results support this. The counter-intuitive result is that road teams with a bye prep disadvantage fared better then road teams overall.
Edit: EdinburghEagle's Way Of Looking At 7 Day Discrepancy Data
In the comments below, EdinburghEagle came up with another way of looking at the data. I like it because it seems to show the the bye prep advantage work for road games as well as away games. It goes like this:
Advantaged Home Wins As A Percentage Of Games Where Home Team Had Advantage AHW = Advantaged Home Wins AHL = Advantaged Home Losses DAW = Disadvantaged Away Wins AHL = DAW AHW/(AHW + AHL) = 100/(100 + 56) = 64.10
So, out of the set of games where the home team had the advantage in prep time, they won 64.10% of the time, a fair amount better then they would have won (57.374%) had the bye prep not been in effect.The analysis runs along similar lines for advantaged away wins:
Advantaged Away Wins As A Percentage Of Games Where Away Team Had Advantage AAW = Advantaged Away Wins AAL = Advantaged Away Losses DHW = Disadvantaged Home Wins AAL = DHW AAW/(AAW + AAL) = 53/(53 + 59) = 47.32
So, out of the set of games where the away team had the advantage in prep time, they won 47.32% of the time, a fair amount better then they would have won (42.626%) had the bye prep not been in effect.
These results make more common sense right?
The Bye Prep Advantage: How Teams Have Fared Historically
Given that the bye prep advantage can give a team an important edge, I thought it would be interesting to see which teams did a good job of taking advantage of it when they had it, and also which teams dealt well with the adversity of being at a bye prep disadvantage:
|7 Day Discrepancy Favorable Games Summary By Team Win Percentage - Top 10 With Ties|
|Team||Count Of Favorable Circumstance Games||Count Of Favorable Circumstance Wins||Win Percentage|
|New England Patriots||11||9||81.818|
|New York Jets||9||7||77.778|
|Green Bay Packers||6||4||66.667|
|New Orleans Saints||9||6||66.667|
|7 Day Discrepancy Adverse Games Summary By Team Win Percentage - Top 10 With Ties|
|Team||Count Of Adverse Circumstance Games||Count Of Adverse Circumstance Wins||Win Percentage|
|Green Bay Packers||8||6||75.000|
|San Diego Chargers||9||5||55.556|
|St. Louis Rams||8||4||50.000|
|New Orleans Saints||6||3||50.000|
|New York Giants||6||3||50.000|
The Bye Prep Advantage & The Eagles 2012 season.
The Eagles have handled teams coming off of their bye pretty well in the last 10 years, finishing 9th overall with a win percentage of 55.556. 4 out of their 5 wins in this circumstance have been road games.
On the other hand, Pittsburgh had been really good at capitalizing on the bye prep advantage finishing in 6th place with a win percentage of 77.778. 1 of their 2 losses came at home.
The Redskins have not been good at capitalizing on the bye prep advantage finishing in 27th place with a win percentage of 37.500. The Lions have been mediocre in this circumstance as well, winning half of their games, with 3 of the wins coming at home.
From where I sit right now, I'm most worried about Pittsburgh. I think its at least a little interesting that the Cantor Gaming NFL Spreads has the Pittsburgh game as only one of 2 the whole season where the Eagles are the underdog. But NFL games rarely work out the way one expects them to during the summer!
GreenInBaltimore Style Points System Rankings.
In GreenInBaltimore's post he devised a point system for figuring out overall discrepancies in prep time by calculating the advantage or disadvantage, for each game, that a given team had in prep time compared to that team's opponent. DerbyGuy wanted to see if there were any instances where a team had less prep time overall then the Eagles will have in 2012. GreenInBaltimore figured out this number at -31.
Well the results are in and -31 is worse then anything I could find! I don't have the results for the rest of the teams from the rest of the 2012 season, but I'd be really surprised if any team was as disadvantaged in prep time as the Eagles are going to be.
|GreenInBaltimore Point System Summary - Top 5 With Ties Worst Disparities|
|Team||Season||Season Total GIB Points|
|San Diego Chargers||2005||-23|
The bye prep advantage lies solely with the home team, doubling the home field advantage. Therefore, I think the NFL schedule makers should avoid this situation, if at all possible, in the future. Edit: As of 5/11/2012 I think that the bye prep advantage confers extra wins to both home and road teams.