Give Me The Rock
this is fantasy basketball
An email discussion between Nels, Erik, and I led to an interesting question: Do the number of transactions a team makes in a fantasy league have an effect them making the playoffs? Or having a better record? Do better teams in a league make more transactions?
With a little help, I pulled the results from 12 different leagues Nels and I were in over the past two years. Now these might not be exactly representative of your average fantasy league because these leagues consist most of the Readers’ Leagues and some experts’ leagues, but it’s the data that we have to work with.
The Ground Rules
A few notes about the process.
1. Transactions include adds, drops, and trades, although we don’t have the information to break them out separately.
2. I looked at where teams finished the regular season (not the playoffs, which are based more on luck) and compared that against the number of transactions a team made over the course of the year.
3. Since leagues have a different number of teams in them, I had to do something to normalize all the leagues so that they could be compared with each other. Basically, a ninth place finish in our GMTR 20-team Monster League is a lot different than a ninth place finish in a 10-team league.* The first solution that came to my head was to use percentiles. No need to bore anyone with calculations, but if you think back to SATs or whatever, a first place team would be at the 99th percentile (meaning 99% of other teams finished lower… go with it) while a last place team would be at the 0th percentile (no teams are lower).
4. All (or nearly all) the leagues were either weekly changes leagues [insert Readers league bitching here] or had a games played cap on them – so this is strictly a streaming-free zone. That is a useful thing, because the topic of streaming is an entirely different question and can be looked at in a separate analysis.
The Good Stuff
I took the results of our 12 leagues and graphed them on a scatterplot, with the total number of transactions made by team on the x-axis and a team’s regular season finish (formatted as a percentile) on the y-axis. I’ve included a trend line in red, which shows how a team’s finish is likely to increase as the number of moves it makes increases.
Table 1: League Finish vs. Number of Transactions Made
Done and done. More transactions = a better team. So says the trend line that increases as the number of moves made does. According that line, if your team makes at least 189 transactions next year, you’re guaranteed a first place finish.
However, there are a couple things that bothered me about the data. First, while many of these leagues were competitive, there were still some teams that made 0 transactions all year. Basically, they were completely inactive. You can see on the graph that teams who make no transactions all year have horrible finishes. While that’s somewhat interesting, it is clouding our analysis of the real question of ACTIVE teams who make more or less transactions.
Also, on the other end of the graph are the people I dubbed “the crazies”.** Most teams in our leagues made somewhere between 0 and 40 transactions in a year. However, there were 12 teams that made over 50 transactions in a single year and the highest number of transactions made by a team was 95.
So, what does the trend line look like if we remove the inactives (classified as teams making less than 5 transactions in a season) and the crazies (50 or more transactions in a season)? Here is the same graph with those two groups broken out separately.
Table 2: League Finish vs. Number of Transactions Made (Split into Groups)
Surprisingly, you get almost exactly the same type of trend (can’t say I was expecting that). Among the people who make between 5 and 40 transactions in a season, the more transactions a team makes the better its league finish was, on average. Just like before.
But when you think about it, 5 transactions over the course of an entire season is still a really small amount. It’s less than 1 transaction a month. What if we only look at teams that are fairly active – teams that make at least 15 transactions in a year (with no cap on the high end this time)?
Table3: League Finish vs. Number of Transactions Made (Teams with 15+ Moves)
We finally see something interesting and a little unexpected going on. Among the teams making 15+ moves, the trend is flat, indicating that transactions stop improving a team’s standing after a certain point (in this case, after 15 moves). Fairly active owners who make 2-3 moves a month do about as well as the people making 30, 40, or even 95 moves a year.
In case you’re interested in a full breakdown of the results, here is a table that shows the average league finish (by percentile and then translated into what it would be in a 14-team league) broken out by the number of moves made. So, teams that made no moves finished in the 25th percentile, which translates into 10.5th place in a 14 team league. That number jumps to 7.8 among teams who made 1-4 moves and 6.4 for teams that made 5-9 moves. Interestingly, the group of teams who finished the best were those who made between 16 and 20 moves a year.
| Moves | Percentile | Avg Finish (in a 14 team league) |
| 0 | 25% | 10.5 |
| 1-4 | 44% | 7.8 |
| 5-10 | 54% | 6.4 |
| 11-15 | 51% | 6.9 |
| 16-20 | 65% | 4.9 |
| 20-40 | 59% | 5.7 |
| 41+ | 62% | 5.3 |
Whatcha talkin bout Willis?
One of the limitations of this study is that we don’t know anything about the makeup of the teams. It could be that those crazy owners making 50 plus moves a season were simply trying to keep bad or injured teams afloat (I bet that has some truth to it). Maybe those teams would have finished much lower in the standings if it wasn’t for their managers deftly making moves week in and week out. And maybe those awesome teams making 16-20 transactions were blessed with number 1 picks in the draft and super healthy teams. It’s certainly possible. But without taking a close look at each of the teams one by one, it’s impossible to say if that’s the case or not.
What we can agree to agree upon for now is that making 15 or more moves appears to put your team in position for a good league finish, while making more than that won’t necessarily improve your chances.
* Another way to do it would have been to look at leagues containing the same number of teams. For example, only look at 14-team leagues where any 5th place finish was equal to any other. But that would have required a lot more leagues than we had access to.
**In the world of statistics, they’re called outliers… but let’s call a spade a spade here.
Subscribe via RSS