Drive Chart Dogs in the GAP matchup grid
One of the best statistical predictors for us over the years has been the drive chart projected scores that come in showing an overlay for the underdog. These prime picks have covered 57% for any level of overlay (even the most modest 1/2 point difference from the actual line), 59% with a 2+ overlay, and 63% with a 4+ overlay.
Last season was not by any means the most successful overall for the 'DC Dogs' but they still managed a 12-8 (60%) mark with the 4+ overlays...and again this represents plays against a neutral consensus line, not the optimal wager.
In other research we've done we have looked at pairing up teams based on a Good-Average-Poor breakout for their records, both in-season and the prior season number (since this can have such a large carryover in people's perceptions of teams). We will revisit this scheme here, using the following numbers for the prior regular season win tally:
"Good" -- 10+ wins
"Average" -- 7 to 9 wins
"Poor" -- 0 to 6 wins
With this in mind we have reviewed the Drive Chart Underdog overlay choices during the regular season matchups (which excludes a healthy run of the DC Dogs in playoff games) in the week 5 through the end of the regular season timeframe.
DC Dogs 'All' Overlays
| Overlay Team |
Opponent |
| Prior Wins |
<7 |
7-9 |
10+ |
| <7 |
52-42 (55%) |
19-20 (49%) |
27-19 (59%) |
| 7-9 |
13-6 (68%) |
27-19 (59%) |
32-18 (64%) |
| 10+ |
2-2 (50%) |
13-9 (59%) |
20-21 (49%) |
So the perhaps to be expected "poor" prior season team facing a "good" prior season team showed some nice returns, but it was the average team from the prior year that was the best overall drive chart dog play, who trounced in particular the poor and good sides of the previous season.
More importantly perhaps is the 4+ overlay territory, and here are the splits for that:
DC Dogs with 4+ Overlays
| Overlay Team |
Opponent |
| Prior Wins |
<7 |
7-9 |
10+ |
| <7 |
26-14 (65%) |
5-8 (38%) |
18-12 (60%) |
| 7-9 |
2-2 (50%) |
9-5 (64%) |
16-8 (67%) |
| 10+ |
2-0 (100%) |
7-2 (78%) |
12-9 (57%) |
With sample size becoming tiny it's questionable to go to this level, but it is comforting to see that all classes fared well with the exception of a poor prior year team as a DC dog pick to an average prior year team.
Casting an eye over the opponent breakouts it tallies up to be 30-16 (65%) when the DC Dog pick is facing a poor prior year team, 21-15 (58%) facing an average team, 46-29 (61%) facing a good prior season team.
|
