crockpot's mlb picks 4/16-4/22 (YTD 30-29, +38.12)

[dankhank]

so what is the edge on a juice discount of five cents?

[Homer]

[ QUOTE ]
so what is the edge on a juice discount of five cents?

[/ QUOTE ]

it depends on the particular lines since 5 cents doesn't always mean the same thing (difference between +100 and -105 is much more than the difference between +300 and +305), but in this case being able to bet at +100 instead of -105 is about a 2-2.3% increase in roi (depending on what he thinks the true line is). i guess if you don't see a response of "awful" by ImaScammer, you can assume i'm right.

[coolmoedee]

Yanks at +168 is really nice and i've bet it large. Go Yanks!

[MrFeelNothin]

All I have is 5dimes and bodog, but bosox have dropped to -155 at bodog while yankees are still +155 at 5dimes. Are there still better lines for the yanks elsewhere?

[TheRover]

You have Bos-155 available. Why aren't you betting it?

[MrFeelNothin]

I did.

Seems like there was an arb opportunity for a while there though. Now bosox are back to -160.

[Domeo]

back to back to back to back

[crockpot]

was out at the wrigley bleachers today. beautiful weather, if a bit sunburn-happy. great game, one unlike any i've ever seen. every burst of offense, up until the last, was met with an almost immediate counterattack, leaving the game tied at 5 different scores. no half-innings with 1 run scored, only one play (solo homer) that scored just a single run.

the fangraphs account provides an accurate portrayal of all the momentum shifts in this game, but doesn't consider the atmosphere of a cubs-cards game.

maybe the most enjoyable game i've ever been to, even though the wrong team won. you definitely had to be there. the worst part is that with the red sox playing the yankees and hitting four straight homers, espn seems to have forgotten this game was ever played.

[Post-Oak]

[ QUOTE ]
crockpot corrected me with a solid explanation, which I appreciate.

[/ QUOTE ]

Actually, the bulk of crockpot's explanation was nonsense. His example is also terrible, and also wrong (actually it's up for a little bit of debate since a true 70 win team will win 95 games less than 1 in 15,000 times, which means you could pretty much ignore the greater EV of the under 100 wins price he mentions).

Only his last sentence was relevant to the case of a baseball game total.

King Yao recently made a post on this forum where he asked people to read a couple of chapters from his upcoming book, and provide criticism. As it happens, one of those chapters concerned comparing the value of baseball totals.

In this chapter he points out that baseball games land on an odd score much more often than an even score. The reason is obvious, since games can not end in a tie. He further points out that the actual total of a game falls on a higher total much less often than a lower total (i.e. games with a total of 7 will land on 7 much more often than a game with a total of 11 will land on 11). This second point is the relevant point crockpoint mentioned in passing, but it is far less important than the first point.

The result of these two factors is that the relationship between totals and prices is not linear, as Thremp pointed out in his usual DBag way (no explanation, just enough to try and make you look bad).

King Yao has access to a database of all MLB games which occurred from 1998-2005, their lines, and the results. He shows how he calculated the following "rule of thumb" values for how valuable a half run would be in terms of price:

Total/Half Run Value
7 29 cents
8 16 cents
9 23 cents
10 14 cents

The half run is worth a lot more for odd numbers. The half run is also worth more the lower the total is (when comparing odd to odd and even to even).

Now these are just the rule of thumb values, and they are less and less accurate as you get away from even money (+100) prices. In the book, he shows how to calculate the value of a half run given two different lines (given his statistics and certain reasonable assumptions).

For example, his database shows that games with totals of 6.5, 7 or 7.5 landed on 7 12.5% of the time. If you assume that 7 is the fair value total, then the game will go under 7 43.75%, land on 7 12.5% and go over 43.75%. That would mean that the following three lines would have equivalent value:

Over 6.5 -149
Over 7 -120
Over 7.5 +109

You can use his data (summarized in charts) along with the assumption that a given total is the fair value total, to compute the relevant value of any two totals/price. If you don't have the time or energy to do the math in each individual case, you can use the "rule of thumb values".

His data does not include games with a total of 11. Since I don't bet baseball totals, I am not going to search for this info on the internet, but I wouldn't be surprised if you could find a "rule of thumb" chart which includes a greater range of totals.

Again, this is all taken from the chapter in King Yao's coming book. From the two chapters I had the privilege of reading, I can tell that this is the type of book I will definitely buy. It looks like it will be full of usefull info (again, I only read the two preview chapters).

[Thremp]

Post-Oak,

Excellent explanation. Senseless shot about the season win totals, as your figures factor perfect information which has been shown time and again to be incorrect.

Senseless plug about Yao's book. I'll likely purchase it as well, but like most sportsbetting books it'll have some semi-pointless info along with the goot.