Punting!? There’s No Punting in Baseball!

Let’s get a disclaimer out of the way; I’m not writing to recommend a punt and I don’t think you should just completely give up on a category. Ok? All I’m saying is if I were to punt, here’s how I’d do it. Punting in fantasy baseball is when you abandon a category. Saves are hard to come by and you might just be able to completely forego accumulating them. Many fantasy managers will draft with the punt in mind straight from the get-go. Some will wait and see how their team is shaping up and will punt a category that is lacking. Surely there has been work determining the value of the punt, Ron Shandler’s forecaster comes to mind. But let’s look at it from a 2021 perspective and whether or not it’s theoretically possible to punt a category and also win your league.

For starters, I’ll be writing with a 10 team (OBP instead of AVG) roto league in mind, but you can make some easy tweaks to this process to match up with your league settings. I have 10 categories and for each category, I could get any number one through ten. If I punted one category and dominated the rest, I’d have a final score of 91 ((9×10)+1). That should be enough to win. In the past five years of my humble little family and friends roto league, the league champion scored an 85 on average. But what if I punted two categories and dominated the rest? Well, simple math here, I would have 82 ((8×10)+2) points and would certainly be in the running. Here are these variations, written out:

1 + 10 + 10 + 10 + 10 + 10 + 10 + 10 + 10 + 10 = 91

1 + 1 + 10 + 10 + 10 + 10 + 10 + 10 + 10 + 10 = 82

So the real question is, how many different ways can I add these 10 numbers (1 through 10) to get a sum of 85 or greater? If we look at this problem with the thinking that it is a selection with repetition, then we can think of it the same way it is written in Combinatorics: A Very Short Introduction:

If we select k items from a set of n objects, and if the selections are ordered and repetition is allowed, then the number of possible selections is n^k.

With this thinking, the numbers 1-10 can be arranged 10 billion different ways (10^10), but that doesn’t help us fantasy baseball managers all that much. In order to see how punting a category affects the other category scores, I decided to run a Monte Carlo Simulation that randomly selects a number 1-10, nine times and then sums the numbers up. I’m using nine because I know one of my scores will have to be a one. Think of it like rolling nine, 10-sided dye. If you roll a total of 84 or greater, you win. The number 84 represents my totals for nine categories and if I add in that extra 1 (from the category I’m punting), I’ll have 85 points and hopefully enough to win my league. I ran this simulation 10 million times and I scored a total of 84 or greater just 23 times. Still thinking of punting? Showing all 23 combinations doesn’t look great, so I’ll just show you a few interesting ones:

Based on the simulation results, the lowest single number of points that I can afford to score in addition to the punted 1, is a 6. Beyond that, the least amount of 10s I can accumulate is five. Punting and doing really well go hand in hand.

Now that you see what it will take to punt, let’s consider what punted category will have the least effect on all the others. In other words, which category correlates the least with all the others. If you punt home runs, your run and RBI categories are going to suffer. Therefore, you would want to punt in a category that is least likely to affect the other categories. To show this, I took all of 2019 stats and was very liberal with batters, requiring only 20 plate appearances to qualify. The same went for pitchers, requiring only 10 innings pitched. I did this because those players affect our fantasy seasons. Think of that late-season call-up that you were holding on to all year. Of course you’re going to start him! Here are the results of the correlations:

I created the total row because I wanted to see which scores have the least impact on all the others. Did we need this fancy math to tell us that the categories to punt (should you choose to do so!) are stolen bases and saves? No. But, it probably does help if you have a lock on saves and are looking for the next best category.

To sum up this wild and crazy experiment, let’s leave it with these bullet points:

• If you choose to score a 1 in a category, be prepared to score a 10 in at least five others.
• The best categories to punt are stolen bases and saves, but Wins and OBP are the third and fourth lowest correlating categories.
• Don’t expect to punt and score anything lower than a 6 in addition to that 1 you’re willing to score.
• Hope all you need is 85 points to win.

*Update – The totals column in the correlations chart was updated to sum the absolute value of the correlations, given that we are just looking to minimize the relationships. This had no effect on the 2 most punt-worthy categories, but it did change the second-best pitching category from WHIP to Wins.

For anyone interest in the python code that generated the Monte Carlo simulation, check out my GitHub page.