The Look Ahead Effect Quantified - NBA

[suicidekings]

http://www.statfox.com/statfoxnews/news~articleid~4907.htm

I was browsing through some online articles and I stumbled on this one from StatFox. Apologies if this topic has been covered in depth in the past (I didn't find anything in my quick search). IMO, it's worth a read.

Essentially, it discusses the validity of the revenge factor and teams looking ahead to future games, and the effect they have on actual ATS performance. The study was based on 3 years of NBA data from 2005-2008.

The results, briefly summarized:

Revenge Games:
The Myth:
Teams perform better in games with the opportunity to get revenge on teams that embarrassed them previously in the same season (winning by 15+ points).

The Trends:
- ATS Rate for teams in revenge games: 46.6%
- ATS Rate for teams in revenge games on the road: 44.7%
- ATS Rate for teams in revenge games on the road when the game right before it was also on the road: 40.3%

Look Ahead Games:
The Myth:
Teams are often assumed to look ahead to upcoming big matchups, leading to lower performance in the game directly before that revenge game (Look Ahead Game).

The Trends:
- ATS Rate in Look Ahead Games for teams with winning records: 43.3%
- ATS Rate in Look Ahead Games for teams with losing records: 48.3%
- ATS Rate in Look Ahead Games preceding revenge games against divisional rivals: 41.1%
- ATS Rate in Look Ahead Games preceding revenge games against non-divisional teams from the same conference: 51.9%
- ATS Rate in Non-Conference games Look Ahead Games: 52.2%
- ATS Rate in Look Ahead Games vs strong opponents: 44.5%
- ATS Rate in Look Ahead Games vs weak opponents before a revenge game against a strong opponent: 37.8%
- ATS Rate in Look Ahead Games on the road before revenge games at home: 35.6%
- When a strong team faces two weak teams in a row, there is no effect on the ATS Rate, but the Over Rate in the first game rises to 56.9%
- ATS Rate in games preceding a 3 day break for the team: 42.1%
- ATS Rate in games preceding a 3 day break for a team when their next game is also at home: 38.1%

Combining Factors (Limited Sample Sizes):
- ATS Rate for strong teams against weak non-conference teams, preceding a game against a strong team: 40.4%
- ATS Rate for strong teams against weak non-conference teams at home, preceding a game against a strong team on the road: 26.3%

[suicidekings]

I've always felt that the notion of revenge games is horribly over-used, and the above findings indicates that the effect is not what people commonly believe. At the same time, the effect of teams looking ahead to big match-ups is shown to be a legitimate trend under certain circumstances.

I haven't actually put the trends to the test yet, but either way, it's an interesting bit of information.

[Peep]

Another thing about "revenge games" is that the team got beat by 15 + for a reason. Maybe they just don't matchup well with the team they want revenge against.

[sofos]

Nice info, Suicide. Do you have sample sizes?

[suicidekings]

Sample sizes aren't given for all the stats, but for those that are, all but a couple of them are based on at least 125 games spanning the 3 seasons studied.

[suicidekings]

Peep: I totally agree. However, I still think the data has some value, particularly in showing trends towards poor ATS performance when playing a weak team before a strong one and when a team plays their last game before a few days break.

i'll start comparing the trends to team-specific data from this season and see if anything pops up as notable.

[pico]

very nice thread. i always wondered about these myths cuz that is all my gambling buddies talks about.

[Wrecktangle]

Old news actually, for both NBA and NFL. In the 80s, a number of folks had drilled out the "revenge/look ahead" factors. Actually, almost every "angle play" falls apart when examined.

Shows the value of maintaining long dbs to check things out.

[Dark Horse]

The definition of 'revenge' in this case is either unorthodox or ignorant. A loss by 15+ points in the long NBA season means nothing. It would be different if a favorite was upset, and met that dog again within a reasonable amount of time. Example: the Suns were upset last night in Sacramento, and play the Kings again within a week. The Suns were embarrassed, the game is still fresh in their minds. I could see 'revenge' there.

But you can't really isolate one motivational angle without weighing it against others. A DD favorite, as the Suns will probably be, almost by definition is under-motivated in the NBA.

[joesdql]

To see how the revenge factor works in 5 point increments use the SDQL:
P:margin<0,-5,-10,-15,-20,-25,-30,-35,-40,-45

http://sportsdatabase.com/nba.py/query?text=P:margin%3C0,-5,-10,-15,-20,-25,-30,-35,-40,-45&sort=query_header

It doesn't look like there is much juice there.

cheers,

Joe

[suicidekings]

Neat website. I like it.

However it does show a poor SU performance in the revenge games, further taking away from the myth that teams will perform better in the game after losing to a rival.