Thursday, 3 October 2019

Basic Baseball and Mattress Mack

Pinnacle tweeted out a link to a year old article on baseball betting, and as is often the case, statements are casually thrown out that really aren't helpful or true.

The articled is titled "Why basic data isn't enough when betting on MLB" which is neither snappy nor accurate. 

In fact, as readers will know, basic publicly available data is quite enough to make a decent profit from this sport as my 2019 season summary showed.

The article contains the suggestion to "always consider the importance of the individual game" which would seem to be superfluous to anyone who takes betting seriously, and this paragraph is problematic:

Baseball is a rotation heavy sport. An MLB team has a 40-man roster from which to select their players (this comes from the active 25-man roster). Those new to betting may note that the starting pitcher is often listed alongside the odds for the game and (while some argue it is no longer as important as it once was) this shows the value that can be placed on specific individuals playing in a given game.
To be clear, it is the 25-man active roster that comes from the 40-man roster, not the other way round, but less pedantically, the suggestion that starting pitcher "is no longer as important as it once was" is palpable nonsense. 

A look at the number of sub -400 lines this season and the names of the starting pitchers will show that the starting pitcher is as important to the win probability, and thus starting prices, as ever.

The very next paragraph is this:
If the game you are looking to bet on doesn’t mean that much to one of the teams, it is likely that this will be reflected in the starting line-up - betting without having an in-depth look at each teams' line-up will cost you dearly.
I'd caution anyone serious about betting not to go looking for a game to bet on, rather the process should be to develop strategies and bet on games, and in markets, that are qualifiers.

The article contains some over-complicated and time consuming filler such as this:
Another factor that bettors need to think about when trying to analyse the two teams competing in a single game is day vs. night. Using the Blue Jays as a point of reference again, they have a sOPS+ (day/night split OPS relative to the league, with 100 being average) of 97/109 for night/day. Again, this is far from an in-depth analysis but just goes to show how spending the time to delve a bit deeper into data can benefit bettors.
Perhaps it's not surprising that no evidence is presented that anything in the article actually led to profits. 

To be fair, when I replied to the Tweet and suggested that over-complication wasn't helpful, and linked my Season Summary post, the author Benjamin Cronin replied positively:
Completely agree. The intention of the article is not to suggest more complex data is required, but that the analysis of the data you use is just as important (which you've done a great job of showing in your article).
Hopefully the constructive criticism above will be taken in good faith also.

There were a couple of other comments regarding the Season Summary, with my old friend A Lucky A Day asking:
The numbers for all my baseball systems come from Killer Sports, and although they do update the lines suggesting they are near to closing prices, I don't believe they are claimed as closing prices. I'm not aware of any sites offering closing prices so Killer Sports may not be perfect, but as perfect is the enemy of good, they are more than acceptable. If anyone has closing prices, feel free to send them my way.  

Barrakuda had this to say:

Other than the All-Star Break adjustment to the Overs Total System which is detailed in the post, the systems were all unchanged. 

Regarding whether all the systems I have ever conceived are shown, the answer is no. When I first first started betting on baseball in 2005, the focus was on in-play trading, and looking at edges from such things as starting prices sticking even after the starting pitcher was out of the game.

After hitting the Premium Charge Threshold in September 2012, it was hard to justify spending the time needed for in-play, and I started looking at a more systematic approach that would only take a few minutes each day. The Premium Charge also had a negative impact on liquidity, so the transition was a no-brainer.

Obviously some ideas were discarded almost straight away after some initial research, and others are proprietary and haven't been discussed on the blog, but the simple Totals Systems, T-Bone, Home Improvement and Rhenium etc. have all been around for a while. And of course I'm always open to new ideas.

Some of the more calendar specific systems for opening week, first week after the All-Star Break etc. were not included in the season summary.

Two systems that I no longer bet are the Mountain High System which after six winning seasons had a losing season in 2016 and was abandoned in 2017 after the poor results continued, i.e. the markets corrected themselves, and the UMPO system for the playoffs which proved to be too volatile and not worth pursuing with such a small sample size (around 30 games) each season. Apparently I didn't even update the numbers for 2018 yet. 

It's important to understand that market inefficiencies shouldn't persist for long, and any system should become worthless sooner or later. One needs to stay on top of systems and trends, one good example being the trend towards more runs meaning that the Overs System needed to be adjusted. Games change and markets change, but with change comes opportunity. 

There may be other examples of ideas in this blog, but honestly after 11 years it's hard to recall every system I may have put out there for consideration. No posts are ever deleted so there's a record of the systems out there.    

The playoffs continued last night with the American League Wild Card game in which the Oakland Athletics were, as they were last season, eliminated - this time by the Tampa Bay Rays.

For the first time in MLB history, four 100-win teams are in the playoffs, including the 107 wins Houston Astros who the Rays now have the pleasure of facing in the ALDS. 

Some of you may have seen articles about the Houston businessman 'Mattress Mack' who is looking to hedge his company's exposure to an Astros win by backing them to win the World Series. 

He 'only' wants $10 million on them, but sports books do not take sides, at least to any significant extent, even if the odds are hugely in their favour:
Sportsbooks maintain a much bigger house edge on futures markets, like the odds to win the World Series, than on single-game wagers. Even so, Curtis said multiple bookmakers have declined to take their action on the Astros. And the other shops willing to take their action so far took much less than they wanted to bet.
The Las Vegas Advisor has an article on this topic from someone with inside knowledge on the efforts to place these bets:
Mack had a standing offer from FanDuel to bet up to $10 million at +220, meaning that every $1 million wagered would return $2,200,000 if the Astros win. At the time, that offer was slightly below the market, and Frank and I set out to do better, at least for a portion of whatever was ultimately bet. What sounded like an easy enough task turned out to be anything but, as the books were a lot more timid to take the action than we expected. Why would the books be reticent? This is a futures bet with a 20%+ house edge cooked into the line. But this is also corporate Las Vegas and department heads are loath to take a chance on putting red ink on the balance sheets. Taking this bet, this good bet, still put them at risk and many weren’t having it.
After a whole lot of haggling, several bets have now been made in Las Vegas at +250, led by South Point, which booked it for $200K. That’s good old-fashioned bookmaking from the likes of Jimmy Vaccaro and Michael Gaughan. Four other Vegas sports books took bets at varying amounts: Caesars, MGM, TI, and Circa. Also, one monster $3.5 million wager at +220 was accepted by DraftKings and placed in Mississippi at the Scarlet Pearl. That’s where things stand right now, with close to $4 million in action. But using a hold ’em poker analogy, this is only pre-flop action; there’s much more to come.

No comments: