Adrian's Friday Thoughts [Feb 20 2009]

[diondublin]

A couple of small bets for tonight:

Princeton (-5) to beat Harvard - a very small bet at most, because home court advantage perhaps isn't worth so much in the Ivy League, but Princeton seem almost worthwhile on the standard notions of Kenpom Pythag rankings and conference record. A clincher for this is that late fouls, if our guys have a small lead only, might allow a free throw or two to cross the number.

If you think it's relevant, Harvard haven't won at Princeton since 1989.


Columbia (money line, 6/4) to beat Brown.

Although Columbia (+4, Boylesports) is a respectable proposition too, I prefer the straight-up proposition for structural reasons. The 6/4 with Bet365 is definitely a better proposition than the Columbia (+3.5) that other bookies offer.

I recommend the straight-up proposition though because of the unlikelihood of the number falling on Brown by precisely 1,2 or 3. This chance is less than that implied by the difference between the 10/11 and 6/4 propositions. The likelihood of these numbers is overestimated by the market because of the possibility of extra late free throw points resulting from fouls when these numbers look possible.

Having said that, Columbia (+4), if available, is not too bad, especially if the game is expected to be low-scoring and hence have a lower standard deviation for expected margin.

As to the justfication, there is no way Brown can be 4-point favourites here in my humble opinion. The truth might be Brown by a little, Columbia by some, or about even, but Brown are not true 4-point favourites on power ratings. They are only 1-7 in conference play whilst Columbia are 5-3.

This leads to a sophisticated argument implying a straight-up play. Although there may only be a smallish chance the line is quite wrong, if it is quite wrong it is almost certainly wrong in over-estimating Brown's chance. It is way easier for me to believe that the proper line is Columbia-favourites-by-2 rather than Brown-favourites-by-10. Yet this purported equality is what one should believe if one favours the point spread proposition.

The principle here is that one must maximise the advantage taken when errors may have been made, and that means taking the odds in this case.


A small bet, then, on Columbia, and a very small bet on Princeton.


Good luck,

Adrian Armstrong

[drklrdbill]

Very solid reasoning and logic. You should be a great addition to this forum. Welcome! Good luck tonight.

[CaptainJack]

Adrian has been posting EXCELLENT comments like these on my free pick posts and I've been telling him to get over here in the forum. I think his analysis of the games is amazing and he makes great points every single day.

Thanks for hopping in here Adrian and look to see more from you!

[diondublin]

[quote=diondublin;4657]

This leads to a sophisticated argument implying a straight-up play. Although there may only be a smallish chance the line is quite wrong, if it is quite wrong it is almost certainly wrong in over-estimating Brown's chance. It is way easier for me to believe that the proper line is Columbia-favourites-by-2 rather than Brown-favourites-by-10. Yet this purported equality is what one should believe if one favours the point spread proposition.

This is slightly wrong; I do apologise. If one favours Columbia (+4) then one de facto does not believe in the equality of the two precise positions above.

The principle here is that one must maximise the advantage taken when errors may have been made, and that means taking the odds in this case.

Although the principle, above, is of course sound. When we are 'bifurcated' either very wrong or very right, theoretically perfectly, winning one-half of the bets and losing one-half; we get expected value 1.91 units-for-2 staked when we play the point spread, but expected value 2.5 units-for-2 staked when playing the odds.

One can only ever lose one's stake unit, but can make many times that if taking odds. Tend to take odds-against on the outsider therefore when result standard deviation expectation is high.


In closing, Columbia won 70-57

[drklrdbill]

Yeah, that was the only game my FadePom betting system lost tonight. 3-1 so far, with just Nevada -1.5 left.