Head to head trends:
Look, but don’t rely

Oct 26, 2004 12:15 AM

Last week we concluded that there was marginal evidence that underdogs off six wins in the last 10 games, would be a solid play. This week we will explain the more significant applications of head to results.

We focused our study on in-season prior meetings, as well as how head-to-head history changed the outlook for "over/under" plays.

First up, how does winning a first in-season game change the second contest? Here are our findings.

”¡ Large favorites laying 7½ points or more after winning the first head-to-head meeting are just 77-104 (43 percent).

”¡ Away favorites that won the first game by 21+ are just 8-16 when laying more than a touchdown in the second matchup.

”¡ Home dogs of 0 to 3 points that won the first contest are 19-9 in the rematch.

The actual won/lost outcome may have less bearing on the upcoming "who will beat the spread" thoughts than which team covered in the first event. So, this time we’ll run a check on results using the past spread outcome of the first meeting.

”¡ Teams covering the first meeting by 1 to 13 points are just 256-313 (45 percent).

”¡ Favorites for the rematch are just 139-190 (42 percent).

”¡ Small favorites (1 to 3 points) in the rematch that covered in the first matchup are 54-81 (40 percent).

”¡ Away teams that covered the first meeting in tight games (+3 to —3) are just 44-71 (38 percent).

In the grand scheme of handicapping information, the pure head to head outcome, doesn’t lend a huge insight into this week’s likelihoods. There are many ways to win and lose a football game. We’re just examining the surface level data if we only use scores.

We also studied the formula relating to "over / under. Knowing the past 10 outcomes gives you next to nothing to go on in predicting the next time two teams square off.

However, averaging the points scored and comparing that with today’s totals line, we found that once the overlay does get to a high level, some predictiveness appears to develop.

In cases where the average points is + or 8 compared to the "over/under" spread on the upcoming matchup, it’s been a nice 59 percent predictor.

We also looked at the in-season H2H effect according to what happens if the same totals result in the rematch? If anything, betting on the opposite totals outcome to the first matchup is the way to go. This is especially true in cases where the first game went over by 14 points or more. The follow-up meeting has gone "under" 83-57 (59 percent).

Bottom line: Head-to-head history should ultimately only be a minor part of your analysis.