MLB and NBA Update (4/19/11)
Update: Tuesday, April 19 at 9:14 PM ET
What a difference a couple of days can make in this industry. While the overall picks performance since the end of the MLB free trial has not lived up to the expectations set by the trial (which would have been essentially impossible to replicate every week as we discussed in last week's blog), the picks (MLB and NBA) followed up on Saturday's weak performance (likely the toughest overall day in the site's young history) with two solid days that leave us profitable (barely) overall for the "subscription" period (since 4/9) and feeling like the start of the NBA Playoffs on Saturday was just as much a fluke as anything else. Furthermore, there are many positive trends that continue to have me optimistic about the future of MLB and NBA picks.
The most important thing that I can get across to all is to consider the long-term nature of our investment strategies - especially with respect to money-lines where wagers are often made on outcomes that are more likely than not to lose. We want to win every game as much if not more so than anyone else and we expect to be profitable each day. But that does not mean that it is likely or even possible to be profitable every day or every week. We happen to have a great track record so far in baseball of profitable weeks (on the whole - including O/U) and far more profitable days than weeks, but that does not mean that anything is technically wrong if we take a loss... The tone of those comments is probably a little stronger/harsher than I prefer. I should note that almost everyone I have ever communicated with through the site does get this concept and that I totally understand human nature associated with reaction to losses as opposed to wins. I am very happy with our reception and the relationship that we/I have built with our users; I just feel compelled to make this point (about "big picture"/long-term thinking) as clearly as possible right of the bat.
Here are the MLB numbers to-date (all assumed to be based on $50 player and -110 with no other plays on that day - profit means more than the percentages):
- Money-line Plays: 54.4% +$170
- O/U Plays: 60% +$652
- Normal+ Plays: 56% +$184
- All Playable Plays: +$822
There is something very interesting to note about the normal+ plays that further speaks to the necessity of a long-term view in baseball wagering. On Monday, April 13, we suffered three brutal losses to top plays that, by themselves, cost us about $200 (for $50 player). We had the OVER (6.5) as a very strong O/U play (60.5%) between Cincinnati and San Diego. With the Reds up 3-2 through 4.5 innings, meaning that any Padres' run would ensure an O/U win, San Diego put the lead-off batter on-base in the 6th, 8th and 9th innings (in the 8th and 9th, the batter got to second with no outs). The combined expected runs scored given those situations for those three innings was 2.8 - where, in each chance, the runner had a greater than 35% chance of eventually scoring (meaning around a 13% chance that the lead-off runner never scored in those innings). The Padres (and Reds) never scored again...
In the A's @ White Sox game that same night, Chicago was up 1-0 in the bottom of the ninth inning with one out and a guy on second base when Juan Pierre dropped Daric Barton's pop-fly in left field. The next batter struck out (which would have been the third out and the game without the error). According to our live win probability calculations, Chicago was 85% likely to win that game before the error and still 67% likely to win the game after it was tied going into the bottom of the ninth. The White Sox lost 2-1 (Juan Pierre made the last out.)...
Then, in a game that I had initially marked as a win when I went to sleep and I had to change a couple days later when going back through the results, the Mariners handed the Blue Jays (+153 on the money-line and a strong play of ours) the most unlikely loss of the season so far. Toronto was up 7-0 in the seventh inning at Safeco Field (one of the least offensive parks in the league) when Seattle (one of the worst offensive teams in the league) scored eight unanswered runs in the final three innings. The Mariners chances of winning that game going into the seventh? 0.2%...
The point is not to whine about our losses. The point is to note that it took some very unlikely scenarios to make that a challenging day, which weighs heavily on the outcome of the week. Had all of those plays gone our way, we'd be up over $450 last week alone. We'll be on the other side of those kind of games as well (one could make the cae that we already were a couple times in the trial period). Fluke situations like that will even themselves out over the course of the season, which just exemplifies the need to view baseball over the long-term. We don't want to lose them and we certainly don't expect to, but we know they will happen.
Anyway, it is obvious to see in the numbers that our O/U performance has been stronger than our ML performance. A great deal of that is related to the topic we just covered about the up-and-down nature of wagering on baseball - particularly with respect to money-lines. Until last night, our ML profitability for the subscription period was greater than our O/U numbers. We lost three of our top four plays and it's the other way around now. The more I think about it though, there may be some logical reasons for this to happen. As we continue to learn more about these teams, we hope and believe that our ML numbers will catch up to the O/U performance to-date. O/U is not as reliant on that because it looks at the game on a more macro/overall level.
MLB Wagering Strategies:
Many have and continue to ask about "optimal" strategies to bankroll management and our official recommendations with respect to baseball - notably with ML vs. O/U and how many plays to make. There are two pretty standard answers to that: 1) Everything is relative to each day's confidence levels and your capability to and comfort in taking on risk and 2) There will never be a perfect answer and it is very difficult to draw conclusions off of just 2.5 weeks of data. However, I am comfortable refining my stance on this a little (which is likely the information most of you are checking in to read)...
We are clearly having success with O/U picks in baseball. This plays like traditional betting without great odds swinging picks in either direction. Totals are also not as reliant on individual data (save for starting pitchers - general lineup strength, starting pitchers and ballparks dictate totals) so year-over-year fluctuation for a few players is not going to impact O/U as much as ML. We are not generating many "normal+" O/U, but there are some thresholds that seem to make sense. Total picks with 56%+ confidence are 31-16 (66%) and +$682. The most important number there is actually that there have been 47 56%+ totals on the year. This is about 2.5 per day since the season started. We don't expect to hit 66% of these picks for the entire year, but we do assume that we will generate greater than 56% accuracy. Playing 2.5 O/U picks a day at around a "normal" bet and hitting 58% of them should still generate over $2,750 in winnings (assuming "normal" play accounts for increase in bankroll - it is +$2,295 just risking exactly $50 on each and if odds average -110). For those who prefer a little more action and a little less overall risk, playing four O/U picks a day (which is exactly how many 55%+ plays there have been) at just less than a normal play and hitting 56.5% overall will provide eerily similar results.
On the ML, I believe in the way that we are attributing value there and that we will be as profitable as we expect in the long-run. I still believe that, since most games are bunched together in our play value recommendations, it is important to play more/all playable games to diversify. Certain thresholds may not make sense (such as anything for which we recommend risking less than $10 - which are down $10 on the season... I'm not sure there is anything to this data yet, but the 100 ML picks we have had with wagers between $11-$35 are +$570. This is where most of our top underdogs fall because it is more difficult to get an underdog into normal+ range unless we actually think it will win more often than not.), but otherwise, playing most games is the right move. Given the performance of other O/U picks and the stated concerns about the nature of ML picks early in the season, the case can be made (and I am essentially making it in far too many words), that it is logical to temper action on Money-Lines until we can prove that ML can match O/U in profitability. As we learn and research more about our numbers, this may naturally happen because our confidence will dip until our performance improves - at which point it will increase again (I always prefer that we do the work for you in any situation like this). But everyone is free to do what he/she wishes with the bankroll and I would find it perfectly reasonable to lower wagers on ML by one-fourth to one-half while playing O/U, until we can prove that the ML picks are as valuable as we say.
And then there are Run-lines... As we discussed last week, subscribers have reported some strong success on the run-line. Since we were not promoting, publishing or tracking info, I have left it out of the official numbers for now. I have asked one of our subscribers to help me track run-lines though and we have uncovered some interesting data. It seems that, at least over the first 2.5 weeks of the season, the likelihood of a one-run game in which the ML underdog lost was far greater than the books expected based on the difference in odds given for dogs vs. odds laid on the +1.5 run-line. This user has looked at alternative lines and +/-2.5 as well, but only recently (and without as much success initially), so I am only comfortable looking at and talking about +/-1.5 lines. In total, playable 1.5-run-lines since the start of the season have hit 69.8% of the time and returned +$1,425 for the $50 player. While I definitely don't think that is sustainable or to be expected, there is a point to be made there...
Generally, our top RL plays in confidence were on the top ML underdogs we were picking. It seems as though the betting public (as indicated by the difference in the lines in the books) undervalues the likelihood of a one-run game. Psychologically, the theory would suggest that, if a bettor is going to wager on a ML underdog in baseball, he/she would prefer to get odds rather than lay them. Most people don't want to lay -180 (bet $180 to win $100) odds on a +120 (bet $100 to win $120) underdog to win or lose by one run, when they can bet that team to win at +120. I get that. If we think there is a good chance that the underdog wins, in baseball, there is also a great chance (often 10-20%) that the ML underdog keeps it within one run. This accomplishes two things: 1) it extends the gap between RL% projected and RL% needed as compared to ML% projected and ML% needed and 2) it makes the play a likely one, meaning that it happens greater than 50% of the time, which is great for the psyche and for those looking for consistent payoff as opposed to searching for big bang-for-the-buck plays.
Money-line favorites can be playable on the run-line as well (38% of the sample I received were RL plays on the ML favorite - which is usually playing -1.5 at +XXX odds), but are not as likely because the public usually bets these lines up a little more looking for positive odds (plus, there is often more "juice" in general given by the books for more exotic plays like these).
So to summarize the notes above, early indications suggest that playing 2-4 O/U a night, ML picks greater than $10 in risk and all playable RL picks (only 2-3 a night - and usually on ML underdogs) seems to be a worthwhile approach.
I don't want to spend too much time on run-lines because they are not "official" publications and we don't even currently provide the best ways to analyze them (please see last week's blog for more info). The most important thing that I can say about them is that we are greatly enhancing our ability to provide you with relevant run-line information. We see too much value in our technology in this respect to leave it the way it is. We have met extensively over the last several days to plan many MLB product changes based on your feedback and this value and will be rolling those out as soon as we can (our lead developer prefers that I not put deadlines to new, unplanned projects in the blog - but he's also the best in the world at what he does, so I don't expect these changes to take too long).
In addition to improving the ability to analyze run-lines (and o/u picks) on the Customizable Predictalator, we will be adding more information for games (starting pitchers we used for that day's projections, upset watch icons, teams noted for "no pick" ML picks, etc.). Please contact us with any additional requests (if you have not done so already). Thank you so much for your help and feedback thus far. It is very important to us to solicit and react to feedback in order to provide our customers with the best possible service. Many of you have taken that to heart and we genuinely appreciate that.
NBA Playoff Performance:
- ATS Plays: 3-4
- O/U Plays: 4-4
It's early. We had an extraordinarily rough start to the NBA playoffs on Saturday and have rebounded about as much as we possibly could have since then. As anyone who has ever watched the NBA Playoffs knows, this is the quintessential "marathon not a sprint" sporting event. I've got far more confidence in our engine and the numbers in general than I did last season (I wanted to wait until after the full package sales period to say that because it's not just a "line" to hook people). We have already seen the impact of our efforts to better account for in-series fluctuation in lines and play.
Also, I'm looking forward to seeing the results of our futures and series odds. This information for the first round is now available to all. In every case (to varying degrees), we found value in the series underdog and in the favorite winning in seven games. It essentially takes three (maybe just two depending on the situation) of these instances to occur to be profitable. Early outcomes suggest that many underdogs have a good chance to stick around for seven games - if not win the series.
As usual, if you have any of your own suggestions about how to improve the site, please do not hesitate to contact us at any time. We respond to every support contact as quickly as we can (usually within a few hours) and are very amenable to suggestions. I firmly believe that open communication with our customers and user feedback is the best way for us to grow and provide the types of products that will maximize the experience for all. Thank you in advance for your suggestions, comments and questions.