The Kaleidoscope Process
The game of fantasy football has come a long way since the days of manually tracking stats and collecting roster moves via phone calls. Rabid fans have continued to evolve the format and come up with creative new approaches. The Fantasy Football Ironman Competition is a great example of this; showcasing all of the different approaches to the game in one. As Commissioner of the FFIC, I'm always on the look out for the latest innovations in fantasy football and this year I came across, what I consider to be, one of the most creative new league formats in several years; Kaleidoscope.
Kaleidoscope is a format dreamed up by John Bosch, who wanted to build a Dynasty league inspired by the classic "Choose Your Own Adventure" books. Instead of holding a traditional or auction draft, Managers pick players for their teams using an ADP generated draft board. This means that instead of picking from the same spot in each round, Managers have the opportunity to move anywhere across a round and pick a player. The league also runs a creative and deep starting roster with 2 QBs + a SuperFlex and 2 TEs. This forces Managers to make some tough choices along the way and that is further complicated by a scoring system that includes PPR, PPFD, and TE bonuses.
Upon hearing about the idea on Twitter I was incredibly excited to join the inaugural season and try out this creative new format. Right off the bat the trading was insane. The league hosts 100 different teams and long before the ADP board was even posted Managers were gambling on grabbing picks in one round or another. I admittedly was very hesitant on making these early trades since I had little context for the value (perhaps a flaw in my game when it comes to Dynasty as I'm always unsure of start-up pick valuations).
The most important part of this start-up was the creation of the ADP board, which requires waiting until there was sufficient data. ADP is often a useful concept when it comes to fantasy football, but it needs context. Things like 'how old are the drafts that are included' and 'how many drafts' will have significant impact on where players fall on an ADP list. If the sample size for ADP is one draft then you may see a player go unusually high because there's that one Manager who REALLY believes this is the year. Having a larger sample size will better line-up players relative to Managers' values. This meant that there was a bit of a wait until the ADP data was settled, but finally the day came where the board was posted!
Going through the process of creating a team was the most fun I have had in fantasy football in a very long time. There was so much strategy and thought that went into my choices that I wanted to share a bit about how I approached it.
Instead of just running through and targeting out positions and players, I decided that the best place to start was in evaluating value on a round by round basis. In doing so you're able to find where you might steal a player who is undervalued based on their ADP. No matter what the format, there are usually a few players who drop round after round due to concerns such as player holdouts, age, and so on. Finding these guys who are likely to produce well above their ADP is generally a great way to approach any draft, but extremely important in a format like this because you'll be able to use that in order to target other positions early. Take for example the Aaron Rodgers situation. Rodgers is annually a top QB, but he's been dropping further and further in drafts due to concerns over whether or not he plays this year. If you decide that Rodgers' value is well above his listed ADP here in the 4th round, then you can opt to pass on QBs in higher rounds knowing that you'll be able to get a good starting QB in the fourth where players in other positions may be weaker. Keeping this in mind I set out to convert this draft board into a value heat map in order to help me visually understand where I might find value opportunities.
During this process I used black spots for those players that are on the draft board, but are currently free agents. While you can of course assign value to a player who isn't on a team (and I'm sure Todd Gurley will upset someone with a late signing for example), I prefer to simply ignore them simply due to the fact that there is so much uncertainty in assessing them without a team for reference. Right away you can see areas where my player values may not match their position on the draft board. In Round 22 for example, there are immediately two players in green that stand out as late value versus one that shows up in red as a guy who I think would have been taken too early in this draft. With this kind of information you can better make player evaluations as you're creating your team.
With each player assigned a value I could then create a scale based on the round as a whole (as illustrated by the bar on the left). Right away I know that Round 28 is one where the players don't have a lot of value and my choices there will likely be incredibly limited. I also made the determination that player values progressed about as expected up until about Round 20. After that point the values become more sporadic, which is to be expected in any start-up draft since this would be the point where teams have filled all their starting positions as well as back-up roles and would now be taking shots on players with upside.
The next thing I wanted to focus on was positional location and density on the draft board. Consider a draft round that is heavily populated by RBs - you'll likely want to be picking a RB in that round because targeting another position means passing on all those RBs. This is especially important with shallow positions such as QB and TE. Because of this, Round 1 and 2 end up being incredibly tricky to navigate although for different reasons. In the first round you're likely to go QB since it is a SuperFlex format; however, the trick here is that this first round eliminates eight QBs. No matter who you pick, you'll be missing out on a significant portion of the top-10.
Round 2 is even more complicated since the positional density is more varied. The first thing that comes to mind is the fact that the TE position is often very top heavy. After the top guys are gone the scoring drops off in a pretty major way. With four of the top TEs appearing in this single round you're almost forced to take one or else you're stuck possibly picking more TEs in later rounds to make up for the lower point variance. This is where the problem of the round's even positional density comes in - by going TE in the second round you're opting to pass on the top-4 WRs and are continuing to pass on the RB position, which was also likely passed up in the first round due to the excessive number of QBs.
Like the first round there were a few others that were heavily populated with a single position; Rounds 7, 9, and 16 all heavily featured WRs while 14 was a RB forward round. It was easy to find positional targets when you consider the density of a position versus the whole round (7 of 10 possible players in Round 30 were WRs so that's a 70% chance I'm taking a player in that position). Additionally, considerations had to be made not just on positions within a round but also on the max density per position. While there were rounds with up to 7 players at the RB/WR positions, the largest population of TEs in a single round was 4. This meant that while Rounds 10 and 11 seem pretty open based on an even distribution of positions, they should actually be considered TE forward rounds since they hit that max density at the TE position. Rounds that hit that max population density at the TE position are going to account for just over 6% of the total number of TEs available on the draft board so skipping them in these rounds thins out your options at the position significantly.
Finally I was starting to pencil in not players, but positional targets for each round. While I ultimately wouldn't fully commit to drafting these positions in these rounds, this information was helpful in order to start mapping out my approach...
Round 1 - QB; Round 2 - TE; Round 7 - WR; Round 8 - WR; Round 9 - WR; Round 10 - TE; Round 11 - TE; Round 12 - WR; Round 14 - RB; Round 16 - WR; Round 20 - RB; Round 25 - WR, Round 26 - WR, Round 28 - WR; Round 30 - WR
Now that I had a breakdown of where each position fell on the draft board I could combine this with my earlier heat map in order to determine where the positional values were versus ADP.
With this information I now had a more specific idea of where the value picks were in the draft and was able to further map out positional targets in each round. Take for example the QB values; it shouldn't be surprising that the value on QB picks drops off significantly around the middle rounds. This is the point where, in a normal draft, all 32 starting QBs would be taken and teams would be grabbing back-ups. Similarly, by breaking values out by position we can now see that those later green spots on the full draft board belong to the TE position. These are likely older guys who don't have a lot of Dynasty value based on factors like age, but end up being solid picks for teams pushing for wins since they would be productive in a Redraft format where longevity from year-to-year isn't important.
All of this work was done before I even started to look at player names. Up until this point these valuations are being made just looking at positional space on the draft board and not yet considering Player A vs. Player B and finally I felt I had enough information to start properly evaluating pick trades (although admittedly I would be in a general disadvantage now versus those who gambled on picks earlier since everyone has a better idea of pick value). Right away I knew that the real winners were those who managed to get extra picks in rounds that would require Managers to make really tough choices based on positional need versus best player. The rounds I marked as high value targets were 2, 4, and 6.
Then there were some rounds I determined were somewhere in the middle; rounds where I might want an additional pick there, but that I wouldn't target as aggressively or give up as much for. These included Rounds 9, 10, 11, and 15.
And finally there were the rounds I knew I wanted little part of and would be happy to include as my trade bait. The majority of these included anything after Round 20 where I felt most of the players in the round had little value or I had little interest in, but I also had little interest in Round 16, which was heavily populated by WRs that I felt had lower values. Additionally, I pegged Round 8 as a great trade bait target since it was early in the draft, held players with good value, but it leaned towards the WR position with guys I was willing to pass on.
Now I'm sure there will be plenty of more experienced Dynasty players who are better than I am that who would say all of this was a waste of time since I'm putting so much value on positions, and that Dynasty is all about just getting the best player available, but this is how I approached things to finally be able to evaluate picks per round. Now this league has a lot of great members in it and even up until the last possible hours of the roster deadline trades were coming in. I won't bore anyone with a breakdown on each offer and counter, but I do want to talk about the one trade that I did end up making during the set-up process.
At this point I had mapped out most of the first ten picks. I knew that I was going QB into TE in the first two rounds, look for value WRs late, and hit double RB in 3 and 4. With the league starting 2 QBs and a SuperFlex I made the selection of QBs a priority. Looking at where QBs fell across the board I really wanted to get three NFL starters solidified even if it meant making sacrifices at other positions in order to do so. This meant that the latest of those three QBs that I wanted to grab was going to be either Derek Carr or Jared Goff in Round 8. At this point I was also stuck over what I wanted to do with Rounds 10 and 11. Both were indicating to me that it would be a good spot to grab another starting TE, but to use both rounds on that position felt like it required me to give up too much by passing on other great players, especially when it seemed like I was pushing WR picks further and further down.
That's about when this trade offer came through. The compromise was that I would be trading down in one of the stronger, earlier rounds in order to move up in two later rounds. I currently had slated Nick Chubb in the 3rd, Derek Carr in the 8th and possible a TE in the 11th. By taking this trade I was able to eliminate the debate over Rounds 10 and 11 since the choice would now be made for me and all I had to do was "downgrade" from Nick Chubb. While I do like Chubb as a RB prospect there are also 5 RBs to choose from in the fourth round and frankly I preferred several of them to RBs in the third. The only real hold-up was giving up Derek Carr in Round 8. It was a simple enough compromise because I was just swapping from Derek Carr to a Round 7 QB, but despite there being three in the round it felt like a forced choice into Daniel Jones. The final decision worked out to basically be giving up Nick Chubb, Derek Carr, and Hunter Henry for Ezekiel Elliott, Daniel Jones, and Jonnu Smith - which I felt pretty good about since it opened up my ability to grab other players that I may have passed on otherwise including Antonio Gibson, Chris Carson, and Tyler Boyd.
With this trade in the books I was finally able to get down to the real grind of building out the team. If you're a fan of logic puzzles or even Sudoku then this type of format might be right for you (or maybe I just like to over think and over strategize). With a copy of the draft board in hand I started to go through and piece in players. I had an idea of how many players I was hoping to have at each position and what positions I might be taking in several of the rounds. From there I was able to start eliminating choices and narrowing down the field. Take Round 30 for example, I had already eliminated two free agent players as choices and knew that 70% of the players were WRs. I started off by looking at the non-WR players that were available to see if there was a chance that a guy there was someone I liked so much that I was willing to take them despite passing on a large selection of WRs. After determining that they weren't any more appealing than the WRs, I was able to start comparing my options on a player by player basis to make my final choice in Tim Patrick. This is how I continued to build out my roster, going round by round and crossing out players I wasn't really interested in as well as slowly targeting out positions until I had finally built the whole team.
I won't spend a lot of time picking through each of my choices, but there were definitely some things of note in my final decisions. As you may immediately see I decided to target out a lot of doubles since this was a best ball format and wouldn't require me to make choices on starters each week. This makes it easier to target down the RB2 for the Chargers for example by grabbing up both Joshua Kelley and Justin Jackson. I also thought there was incredible value in Allen Lazard at 21 and Marquez Valdes-Scantling at 27. In my opinion Aaron Rodgers holds little leverage in Green Bay and if he doesn't hold out all year then one of these guys will be the #2 target on a team that is pretty pass friendly.
The league is just getting started, but I already love this format. There are so many levels to the strategy here that I just don't have time to get into, most important of which is a limited free agency. The combination of larger starting position settings, bonus scoring, and so on really do go a long way to create a level of strategy for this league that I think is exactly what hardcore fantasy football players want. I'm definitely looking forward to seeing how this all plays out and what choices my competitors made!