The inspiration for this article came from a similar 538 article a few years back on the Patriots. The first thing we might want to ask is, “How many years does a team need to do well to be considered a dynasty?” Should we just look at 4-year periods since that’s the max years a student can be on a team? Or should we just look at single year periods since every game is so different? A main idea from the 538 article is that it’s kind of silly to ask such questions, as a true dynasty should clearly stand above the field no matter how you chop up the data. I was originally planning to use a very similar visualization method as theirs, more in line with what I made when talking about dynasties back in this thread. However, there are a pair of very notable differences between FRC dynasties and NFL dynasties which I will expand on below. Like a lot of my projects, I’ll be using my Elo ratings as a baseline “team strength” statistic, because it’s very flexible and it’s the best general predictor of competitive success that I’m aware of.
Directly Comparing Team Strengths Between Years
The first distinction is that it’s not nearly as fair to directly compare years in FRC as in the National Football League (NFL) because some games are just naturally prone to creating more powerhouses than others. Whereas taking a raw Elo average works well over any period in the NFL since football in 2018 is still 90+% of the same game as football in 1968, raw Elo doesn’t make as much sense to compare between years in FRC, where even consecutive games are vastly different. Maybe over 5+ year periods, these game differences balance out, but that’s not obvious to me.
So what can we do about this? My approach will be to add a normalization factor based on the standard deviation of Elos for each year. If a year has a higher standard deviation of Elos, it is “easier” to be dominant relative to average teams in that year, while the opposite is true for years with low Elo standard deviations. For the curious, here are the Elo standard deviations for each year:
Since we are looking at each team’s full season performance, I am opting to use each team’s season average Elo, instead of max or end of season. As an example of this normalization, let’s look at 1114 in 2010. Their raw season-long average Elo (as in, the average of their Elos immediately after each of the matches they participated in) was 2083, and the standard deviation of average Elos in 2010 was 72. So their normalized score for the season was (2083-1500)/72=8, which means they were 8 FULL STANDARD DEVIATIONS better than the average (1500 Elo) team that year, which is just absurd.
Visualizing the Results
Another big distinction to the NFL is that FRC’s history is pretty short. We only have reasonable match data going back to 2002, while the NFL has decades. The big plus side to this is that we can still show who was the best team according to this methodology in every year range on a single chart. So let’s do that. Below you can see a graph of the “best” team over each year range 2002-2018. How to interpret the graph is the following: For the question, who is the best team between year A and year B, look at row A and col B on the table. The team listed there is the team that had the highest average normalized rating between the listed years.
What do we see in this graph? A whole lot of Simbot red. Ignoring the 2002-2004 years for which the data is sketchy and the games are still 2v2, there are roughly 3 notable dynasties in FRC. The 233 era, the 1114 era, and the 254 era (with some 2056 sprinkled in there as well). The 1114 era easily stands above the others though. Counting up entries on this chart, I’ve created a new metric I’m calling the “dynasty score”, here are the results:
Really no matter how you slice it, 1114 dominates, which is why I felt fine making my title so definitive. Now this doesn’t mean 1114 will reign forever, if 254 keeps up their pace a few more years then they’ll have a reasonable claim to the greatest dynasty, but if you think 254 is the greatest dynasty already because they’ve been the best on the 2012-2018 range, take a pause and remember that 1114 was the best over 8 such 7-year ranges. Also, 1114’s biggest flaw is honestly that they did not exist in 2002. If they had just existed in 2002, even if they were the literal worst team in FRC by Elo that year, they would STILL earn the rights to the 2002-2010 all the way through the 2002-2018 titles.
A Couple More Charts for Fun
Just because these graphs are so pretty to look at, I’ve made a couple more. First we have the second best team by year range:
It doesn’t really make sense unless used in tandem with the first chart, but it’s much more colorful.
Here’s Minnesota since 2007:
And here’s New York:
You can find charts for all regions here (credit to pchild for the location data), but unfortunately most regions won’t have any pretty colors. You can always add your own colors using the conditional formatting tool. Two small notes on the data, teams who were not active in 2018 will not appear in these charts, and if you’re in a young region, things will be non-sensical in the years before your first team was created.
Anyway, hope you all find these charts as interesting as I do, feel free to make your own charts for all of FRC or your home region and share them (ideally before studying mine so I don’t bias you). The results I get algorithmically should be close to your intuitive understanding of “dynasties”, but I’m always interested in seeing differences.
Until next time,