I seriously don't know why I bother, and I know it will not even be given an ounce of thought due to the arrogance/ignorance of PM, but:
* Try devising for yourself a fairly simple metric, let's call it "Expected Wins".
* Work out each horses (%) chance to win by dividing their tote odds (or fixed) by one /1 - the sum of this calculation will contain the WSP's over-round.
* Convert these percentages into a 100% market, giving you each horse's real chance (less the juice) if the market is true.
* Sum these XWin figures across an entire card, using whatever sized buckets/classifiactions you deem fit.
Fav/2nd Fav - 18x runners - 3.74 XWins - 6 actual wins
3rd/4th/5th Fav - 27x runners - 2.96 XWins - 2 actual wins
6th Fav onwards - 48x runners - 2.30 XWins - 1 actual win
18+27+48 = 93 runners
3.74 + 2.96 + 2.30 = 9 races, also 6+2+1 = 9 races
Extrapolate this over a decent sample (I know you won't) to draw accurate, verifiable an meaningful conclusions. You can then filter by state, by region, by track, by field size - and your arguments/hypotheses may be reviewed with more than the skerrick of care that they currently deserve.
Today, the market said the favorites should win 2.18 races - they over-performed this figure and won 5.
Today, the market said the top two favorites should win 3.74 races - they over-performed this figure and won 6.
Today, the market said the top three favorites should win 4.94 races - they over-performed this figure and won 7.
Today, the market said the top four favorites should win 5.87 races - they over-performed this figure and won 8.
Please don't let your shortcomings (lies) get in the way of a good story (facts, backed by data/evidence).