Week 3: What Statistics are the Best Fantasy Teams the Best At?

 

Paul moved his son, Andrew, to Seattle this week for Andrew’s new job at Boeing. While waiting for Andrew on his cross-county road trip and for the moving truck to arrive, Paul jaunted to Mt Rainier and this gem known as Chambers Bay, home of the 2015 US Open. Dustin Johnson had some opinions.

2022 Year to Date Power Ranks
TOTAL			HITTING			PITCHING				Manager	Team
1	Arthur	3.79	1	Cory	4.25	1	Arthur	3.00		Michael	G and RE
2	Carl	4.64	2	Max	4.38	2	Brian/Josh	3.17		Dave	I Hate Fantasy
3	Dean	4.71	3	Arthur	4.38	3	Dave	3.50		Keith	Bourbon Street Blues
4	Brian/Josh	5.21	4	Michael	4.63	4	Carl	4.00		Dean	Benchwarmers
5	Keith	5.71	5	Dean	4.75	5	Dean	4.67		Arthur	[ALL CAPS TEAM]
6	Cory	5.71	6	Carl	5.13	6	Paul	5.33		Carl	Boston Running Sox
7	Max	5.79	7	Keith	5.38	7	Keith	6.17		Cory	Hebrew Nationals
8	Paul	5.93	8	Paul	6.38	8	Cory	7.67		Max	Pancake Nips
9	Michael	6.14	9	Brian/Josh	6.75	9	Max	7.67		Paul	2nd Act
10	Dave	6.14	10	Dave	8.13	10	Michael	8.17		BJ	Smoak That Ish

We have a breakout! Arthur has made an early move this year and he’s pulled away from the pack after week 3. Behind the bats of CJ Cron, Aaron Judge, and Juan Soto and the arms of Kyle Wright, Logan Gilbert, and Max Scherzer, Arthur is looking very strong. In other news Paul made a bounce off the bottom and is back on track. From the top BJ, Carl, and Dean have all had a couple of tough weeks in a row.

This week we’re going to look at what statistics have led to the best production for their fantasy teams over the years, but before we get all the way into it, there were a few weekly observations I wanted to talk about.

The league leading batting average this week was .266, the high-water mark in Home Runs was 12. Meanwhile two teams had ERAs under 2.00 over a combined 160 innings. Adjust your mental perspective on what good numbers are right now. Hit .240? You might win the week. Pitch a 3.5 ERA? You’re behind the curve.

In the weekly matchups, we actually had some pitching streaming this week. Carl had a big Net Win lead on Dave late in the week, but Dave worked his way back and ended up winning the category. There is some decent starting pitching on the wire, a thought that is not independent of the earlier point about the state of the league run production environment. Beyond a top tier of pitching (40 Starting Pitchers maybe?), the talent level is relatively flat all the way to SP100 or so; therefore, with us owning around 80 to 90 Starting Pitchers, there is bound to be some talent on the free agent wire for good matchups. Stream at your own risk.

OK, let’s talk about some statistics. First off, a few caveats. One, these numbers are entirely retrospective in nature with data from 2012 to 2021. I’m drawing conclusions and projecting them for how they apply moving forward, but things change, the game changes. What was true about team makeup one year might not be directly applicable the following years. Next, these are correlations I’m attempting to show, but they aren’t necessarily causations (AKA, correlation is not causation). For example, I’ll be showing below that having a good fantasy team hitting K/BB rate is a good thing, but if those good fantasy teams were very good at other things as well (like they led the league in Home Runs), then them performing well at K/BB rate and also having a good hitting power rank could be a coincidence (unless you think you need to be good at K/BB rates to be good at HRs, but I digress). Nonetheless, fluctuation in other categories is likely to ebb and flow, therefore if we can get a big enough sample size, this exercise is likely to be helpful.

For the first one of these, I’ll show all the work, then I’ll just publish results to get through these faster.

Let’s talk hitting.

We’ve had one major rule change in our league’s hitting statistics, starting in 2019 we removed the Hit stat as a tracked fantasy category and added SLG. Understandably, that changed what made a team good at the hitting statistics. Throughout this section we’ll call this the H era (2012-2018) and the SLG era (2019-2021 for the purposes of this blog).

During the H era, we were essentially counting hits twice, both as a counting stat and as a part of the AVG formula. Predictably, this meant that if a team was good at AVG, they were also probably good at Hits, and that was two of the eight hitting categories. For perspective, here is what each team’s AVG vs hitting power rank looked like during the H era, by year’s end AVG and Hitting Power Rank:

What you see here shows that a high batting average led to a low numerical power rank, which we all know means good things. The trend line with this shows the downward slope and has an R-squared value of 0.421. R-squared is a measure of how far from the trendline all of the data points are. R-squared values can be between 0 and 1 and the higher the number the better correlation between the x- and the y- axis. A 0.421 value is a relatively strong correlation for the type of data we are talking about: highly independent data (one team’s AVG and almost no correlation on another teams AVG). What this means is that teams that were good at AVG during this time, were also good at the Hitting Power Ranks (AKA were the best hitting teams).

What other data was looked at?

Statistic	Era	R-squared
At Bats	H era	0.315
At Bats	SLG era	0.293
K/BB	H era	0.263
K/BB	SLG era	0.407
K/BB	All Years (2012-2021)	0.283
SLG	SLG era	0.346
HR	H era	0.123
HR	SLG era	0.484
HR	All Years (2012-2021)	0.136

 

On the hitting side, I wanted to see if At Bats mattered. What we see is that there’s very little correlation between the number of at bats that a team has and it’s success in the power ranks. This was the case during the H era, and is even mors so the case now during the SLG era.

Conversely, we see a relatively strong correlation between a team’s K/BB rate and their hitting power rank during the SLG era. If a team had a good K/BB rate (don’t strike out much and walk a lot), they have performed better over the years.

A team’s HR total appears to be more indicative of success than their SLG. So if you’re looking to find a good player for your team, find ones that you think will hit a lot of Home Runs or have a good K/BB rate, their SLG or getting full time at bats isn’t as helpful

OK now let’s look at pitching.

We’ve had one major rule change there too over the years (before this year’s change, of course). We replaced the Earned Runs category with Quality Starts. This rewarded teams for starting pitching and not avoiding earned runs. I’m calling these the ER era (2012-2018) and the QS era (2019-2021). So how did various teams’ statistics compare to their pitching power ranks over the years?

Statistic	Era	R-squared
Earned Run Average	ER era	0.497
Earned Run Average	QS era	0.276
Innings Pitched	ER era	0.062
Innings Pitched	QS era	0.231
QS	QS era	0.336
WHIP	ER era	0.614
WHIP	QS era	0.318
WHIP	All Years (2012-2021)	0.525

A few things surprised me here. I was expecting the Innings Pitched correlation to be higher. I would explain this to the streaming impact we’ve seen over the years, lower tiered starting pitchers just don’t help your team as much as they hurt it. I’d expect this impact even more now with the Net Wins category replacing Wins. Quality Starts has the strongest correlation to Power Rank success in the QS era among statistics looked at, but it’s not a very strong correlation. WHIP is very close to it in terms of it’s R-squared value. The conclusion here is that there doesn’t appear to be a great single indicator of fantasy team success on the pitching side.

So how are things looking this year?

Well, Max leads the league in home runs and is second in the hitting power ranks. Dean leads the league in hitting K/BB rate but sits fifth in hitting power ranks. On the pitching side, Dean also leads the league in QS while Dave leads in WHIP. They sit in fifth and third respectively in the pitching power ranks. We have a long way to go on the season, but I will be keeping track of some of these things to see if these trends continue. Go get some Home Runs!