Meta-Gabeing: Making The Percentage Play
October 26, 2012 | Crazy Gabey
Welcome back to another week of Meta-Gabeing. If you haven’t read last week’s article yet, I highly recommend doing so before proceeding. The concept we’ll discuss today builds directly upon the foundation we laid last week, and will give contextualization to the topic. You can find it here.
I’d also like put an offer out to anyone who finds themselves in a sub-optimal situation. Post a link to your league’s main page (if the league is publicly viewable of course.) Or copy/paste your roster, the roster of each of your leaguemates, and any notable players on your waiver wire. I’ll do my best to help as many people as possible, and I’ll select one scenario to feature each week as a way of practically applying the topics we discuss. I’ve always believed hands-on learning to be a vital component of strategy-based gaming, and I think we could all benefit, myself included, in getting our hands dirty.
With that out-of-the-way, let’s move along.
Making The Percentage Play
A distinct aspect of all strategy based gaming is that percentages are in a constant state of ebb and flow, and can be used to help us differentiate lines of play from one another. Sometimes the percentages are prominent, as is the case in poker. Often though, it is a subtle, unconscious evaluation arrived at through a mix of recall (which we’ll discuss in detail next week) and current information. For example, you might have known off the top of your head, that after week 2 trading for Wes Welker was a good idea. Current trends showed he and Julian Edelman were splitting time, and that Gronkowski and Hernandez were becoming offensive focal points. Recall told us that Welker is a top-10 talent in an explosive offense, and that given his previous body of work, if the Patriots wanted to win they would quickly change tunes in regard to his playing time. By weighing the two against each other, you knew that buying low on Welker was less risk/reward and more of a sure-fire return on investment.
In poker, there are times when a certain play is clearly preferable, and almost obligatory given the odds the pot is offering, how an opponent is playing, and so on. If there is $100 in a pot at the end of a hand you are almost obligated to call a $20 bet if you are holding any strength at all. Of course you may choose to raise depending on the opponent, but the point is you aren’t giving up on a pot like that if there is a reasonable chance you may be holding the best hand. In fact, this situation is so favorable to you that you only need to win 21% of the time to show a profit. To cast it in a different light, you can actually lose 79% of the time and still be a winner! We’re planning on winning more often than the bare minimum, so you can easily see why this sort of opportunity is the kind professional poker players make their rent off of.
Fantasy sports are clearly more fluid and situational, but the basic tenets apply. One of the most overwhelming tasks for a novice is trading. There’s a certain fear grounded in the enormity of it all, and it becomes extremely difficult to accurately evaluate a trade offer. One of the best ways to nullify this is to assign percentages to the pieces you are dealing with. It’s obviously impossible to assign a “hard” percentage to something with as much variance as living, breathing human being, but it’s helpful in making sure you get the deal you want.
For example, let’s say you have Arian Foster, a few solid mid-round guys, and little else in the way of studs. You’ve been decimated by injuries, a few high draft picks haven’t panned out, and you’re currently sitting at 3-4; very close to fading out of the playoff hunt. You clearly need to make a move, but where do you start? The obvious answer is to find an owner with a surplus in your areas of need who conversely needs whatever you might have in abundance. This is the optimal scenario but it often won’t align that way. Instead you’ll need to look for some players to buy low on, an owner who overvalues getting a guy like Foster, or someone who can afford to lose value in order to upgrade their RB corp.
Now comes the fun part.
Let’s start by assessing your current status. All things being equal and realistic, you might have a 15% chance of finishing 1st or 2nd this season. The top teams will be sitting at a significantly higher percentage than this. Your goal is to get as close as possible to those teams, and pass them if you can. If you sit still, hope your guys step up, and ride Foster until the end, sure, you may win once out of 20 times with a couple 2nd places, but this is a long-term losing proposition and an example of playing to not lose. So find that guy who really wants Foster and maybe even overvalues him a bit, if that’s possible. Let’s say he’s got Darren McFadden, Julio Jones, another stud or two, and some mid-range guys. What are the chances DMC starts clicking in the second half? Every owner will have a different answer, but what’s important is that you come to an informed number based off your research, instinct, and recall. Personally, I’d say 40%. He’s shown elite level talent, but has a porous offensive line and a surrounding cast that just can’t seem to click and avoid injuries. What are the chances Julio Jones maintains his current level? Exceeds it? Regresses? I’d give him maybe 15% to “go off” in the second half, 70% to maintain, and maybe 15% that his numbers dip. Let’s say you load the back end of this deal with someone like Mikel Leshoure, giving away Josh Gordon and Steven Jackson. I’d say Gordon has a long shot at continuing his torrid pace, maybe 15%, and I’d reasonably trust Leshoure, say 65%, to continue his solid, albeit uninspiring production. On the surface Foster/Gordon/Jackson for McFadden/Jones/Leshoure seems like a fairly even proposition, but it’s a deal loaded with upside for you. If things work out in your favor, Julio stays consistent, Leshoure continues to be workhorse-like, and DMC turns things around, you’ve probably just increased your chances of placing to something like 30%. If an injury occurs or your guys underperform, take solace in knowing your chances of placing weren’t much worse than if you had sat idle. On the flip side, what if DMC turns his game around and Jones or Leshoure get hot? You’ve probably now increased your chances of placing to something along the lines of 50%, representing a significant upgrade.
To emphasize, let’s take a standard 10-team league with a $10 “participation” fee from each of the members. Assuming a standard payout of 30 Monopoly bucks to 2nd and 70 Monopoly bucks to 1st, let’s see where your fake profit margin lies. We’ll say 15% of the time you’ll place. Given twenty entries you’ll pay $200 in “participation” fees. This means that you’ll make 1st and 2nd approximately 1.5 times apiece for a grand total of 150 Monopoly bucks or a net loss of 50.
Now let’s look at the same scenario given twenty $10 entries and the same prize structure with our team after the trade. We’ll say that 60% of the time our chances remain at 15% due to DMC continuing to disappoint, a key injury, etc. Our profit stays the same in those 12 contests, so $150x.6=$90. We’ll then say 30% of the time we have a 30% chance of placing 1st or 2nd when DMC turns it on. This means that 30% of the time we will have three 1st and three 2nd place finishes so $300x.3=$90. Then we’ll take the final scenario in which the stars align and the upside shines through, giving you a 10% chance of making 1st or 2nd 50% of the time. This means five 1st and five 2nd places 10% of the time so $500x.1=$50. A total return of $230 from our $200 investment.
As you can see it adds up over time, literally. Notice we were still at a $20 loss before adding in the small percentage that represents the upside of our best case scenario, and therein lies the specific reason we must play to win by making bold moves with high risk upside potential. The 10% of the time where it all comes together is the key to turning losing situations into long term winners. Conducting business as usual will never put you in a position where you can be rewarded by a windfall of upside. It’s how you can continue to play to win even when the cards don’t fall your way.
A few last things I’d like to talk about as we wrap up our talk on percentages.
If you have multiple options for your last few roster spots each week, percentages can be very useful in setting your lineup. What chance do you have of defeating your opponent? Is his team significantly better than yours? If so you can increase your percentage of winning the matchup by playing “home run” guys like Vincent Jackson or Michael Vick over other, safer options. The times they don’t “go off” are eased by the fact that you probably weren’t winning even if you had chosen your more consistent players. Conversely, if you know your team is better than your opponents by a fair margin, it’s probably better to play it safe and take your 8-12 points from Steven Jackson than it is to start Shonn Greene and hope for another anomaly.
Finally, I’d like to note that shopping around with trade offers for all the buy low targets is almost always the correct percentage play. I mentioned earlier in the article that unlike poker, fantasy sports are a much more fluid and unpredictable endeavor. This makes for very few instances where there are cut and dry “correct” plays because so much of a player’s value is an inference drawn by the individual owner. Buy low guys are different because you never know just how low an owner is willing to go when he’s frustrated with a Jeremy Maclin after week 5. Perhaps he sells for dimes on the dollar and even if Maclin doesn’t pan out you’re barely worse for wear. I’m not advocating that you acquire every underperformer, but throwing out some feelers is a safe way to improve your team.
Thanks again for reading folks! There’s so much more that could be said on this subject so if anyone has questions I’d love to hash it out in the comments. Come back next week when we discuss how to appropriately use recall memory when dealing with information overload.
Follow me on Twitter @CrazyGabey.