Strikeout

Danny Murphy is back, missing the point of a well known saying in baseball:

"On any day any team can beat any other".

Yes but so what? You can say the same about any sport, in fact that's why we watch it. It really is the most ridiculous platitude.

The question is not can team A beat team B but are the odds for doing so better than the bookmaker's assessment.

Yes, taken literally Tahiti could beat Spain at football, but in baseball, MLB teams do not start at 1.0x.

On any given day anything can happen. That's one of the rare true clichés about baseball. On any given day any team can beat any other team 11-0. The same does not happen in any other major sport.

The point I am making Danny is that the shortest of favourites in baseball is priced longer than you would find in other sport. I'm sorry you missed the point, but you cannot say the same about any sport. That really is the most ridiculous comment! We watched Spain v Tahiti to see how many Spain would win by, not to see who would win. Just this week Akurba FC COULD have beaten Plateau United Feeders but they happened to lose 0-79. Babayaro FC COULD have beaten Police Machine but unluckily lost 0-67*

Thanks for the lesson on value though. That'll be very useful as I try to get my head around sports betting.

Fizzer555 knows his baseball, and actually contributed something sensible:

Ed Smith in his new book "Luck' (well worth a read) describes how some sports are similar to Chess (like Tennis) where the better player/team usually wins whilst other are more like Backgammon (especially Baseball) where there is a much higher level of chance involved but where skill wins through over time.

I know from my own betting I am much better at baseball than sports such as tennis. First of all, I enjoy baseball more, I get into the detail of it more, and I like the mix of skill and luck involved and accept it in my betting patterns.

Re divisional away dogs, I think there is a slight benefit vs non-divisional away dogs but, as you suggest, it needs other filters before it is profitable.

Running the numbers from 1999 onwards (not a best fit start point, just when I have data from) the win-loss results are:-

All div away dogs (W-L) 4059-5609 42.0%
All non div away dogs 5198-7752 40.1%

A lot of that difference, however, disappears when you look by price band

Price band +100 to+125
Div away dogs 2059-2466 45.5%
Non-div away dogs 1380-1682 45.1%
Price band +126 or more
Div away dogs 2000-3413 38.9%
Non-div away dogs 3818-6070 38.6%

Some good stuff in there - note to self "check out the book mentioned" - and totally agree that baseball is a fascinating game once you get into it, and it is precisely the relatively large element of luck that makes it a more even game on any day, but in the long run, the cream rises to the top.

Danny's continued his comment with:

Looking at Pete's post to me he seems in grave danger of falling into the data mining/backfitting trap.

It would be interesting to compare Pete's results to the PECOTA 2013 season forecasts, it might give some guidance through the fog.

Pete - do you know a lot about Baseball or did you get interested thinking you could make money from betting on the games?

I have an explanation for the triple XXX - Pete must have been using Bingo and Adster's 'draw inflation tool'.

To which always the gentleman Peter replied:

Danny,

Thank you for your input it is very much appreciated.

It would be fair to say that I am no baseball expert. My interest in the sport came about a few years ago when I read on this blog that any team could beat any other team.

As I enjoy building and testing systems I then went about compiling my own simple rating system and I came to the conclusion that there was simply no real value to be had backing the short priced home favourites.

Over the coming years this then became more sophisticated to the point where I now use a reliable rating system against various US bookmakers odds. Over the past few seasons simply opposing the home favourite that showed a negative rating has yielded some very good profits.

However even this is now showing a loss of around 20 pts to level stakes.

During last season various rules were developed and I am very wary of your comment regarding Data mining / Back fitting trap and I am just wondering whether some of the rules I have been utilising may have fallen into this trap.

Incidentally where we have backed the strong odds on favourite who is negatively rated away from home this has shown a good profit and I believe the reason for the disparity will be the superior quality pitcher employed.

I need to take a step back over the break next week and look at this all logically because I still believe with a disciplined common sense approach there is good money to be made from this sport. I am just confused by this run which in truth is freakish.

Regarding the XX Draws for some reason I have in my mind the xXx logo of the Vin Diesel (pictured), Ice Cube Films.

With close to 2,500 regular season games each year, with an abundance of statistics from each, there is certainly plenty of data to sift through. Compare this number with the 380 EPL games each year, or the NFL's 256 regular season games.

* There are reports that suspicious betting patterns on these games were observed. Surely a team called Police Machine couldn't be involved in anything untoward? Incidentally, Ian Erskine tipped these two as lay the draw selections; my only regret was locking in profits at 1-0 in both matches. I feel like I gave some value away. Oh well - 72 goals in the second half takes some doing. Neither game qualified as an XX Draw selection.