First Omni 2021 results, plus a look back at drafts
Reviewing some ADP data, plus Champions League, NCAAB, golf results
It’s been a fun first month or so as we get underway for the Omnifantasy 2021 season, and there’s plenty to talk about. Kevin Zatloukal, the man who built the site, has pulled Average Draft Positions from last month and even created what an optimal draft would have looked like based on ADP and the expected points values from mid-February. We also have results to discuss from the Champions League, college basketball, and the first golf tournament, the PLAYERS.
Champions League upset
A lot of the Omnifantasy drafting this year took place after the first legs of the Round of 16, so we saw some changes in ADP as a result. The biggest shifts included Paris Saint-Germain and Barcelona, as PSG destroyed Barca 4-1 in the first leg to all but decide that matchup.
One first leg result that didn’t create a major shift was Porto winning 2-1 against Juventus. Juve had scored an away goal, which ultimately becomes the tiebreaker if the aggregate scoring from the two matches is even, which is to say that if Juventus won the second leg 1-0 at home, their one away goal would have moved them forward with an aggregate score of 2-2. I suspect that was one reason, as well as only being down a goal to a presumably inferior side, there was still optimism they could move on.
Instead, Porto pulled off the biggest upset of the Round of 16 in an exciting second leg. Rather than getting shut out, Porto matched Juve’s away goal in just the 19th minute on a penalty kick. Juventus managed to score twice, though, and force extra time as the two sides were level at 3-3 aggregate with one away goal each. The contest looked like it was potentially headed for penalty kicks until Porto scored again with about five minutes left in extra time. Juventus matched them two minutes later, but that was ultimately futile, as Porto’s 3-2 loss in the second leg meant it was they who advanced on the tiebreaker — their two away goals moved them through despite a 4-4 aggregate scoreline.
That means Juventus finished with zero Omnifantasy points with an ADP of pick 59.2 in the data we’ll look at below. Porto locked in 20 despite going undrafted in many leagues. Along with Porto, seven other clubs moved forward and claimed their 20 Omnifantasy points. Wins in the next round will move the final four teams to 30 Omnifantasy points, while the final two teams will ultimately finish with 80 (for the champions) and 50 (for the runners-up), as in every sport.
One quirk of the Champions League is that in each round, they redraw matchups. That introduces a random element that creates easier or more difficult paths for individual teams that are not known in advance, which has significant effects on Omnifantasy scoring. After PSG dominated Barcelona in the first leg, PSG’s odds to win the whole thing skyrocketed. All they’ve done since is finish Barcelona off, but they wound up with a terrible draw in the quarterfinals, both for them and another favorite.
PSG will face Bayern Munich on April 7 and then again on April 13, as the first and second legs of the upcoming rounds are scheduled much closer together. That has negatively impacted the odds for both clubs, with Bayern still coming in as the second favorite behind Manchester City but at longer odds, and PSG falling to fifth of the eight remaining clubs. With a favorable draw, PSG might have looked like a good bet to reach the final and net either 80 or 50 Omni points; instead, they are now underdogs to reach 30.
English side Chelsea, previously looked at as a mid-level option, got the favorable draw with Porto in the quarterfinals. Their odds to win the whole thing consequently jumped to +450, third best and just behind Bayern’s +400. Chelsea leapfrogged PSG (+700) and also Liverpool (+600), who drew Real Madrid (+1000) in the quarters. Man City were the other “lucky” club and are now the clear favorite (+185), as they drew Borussia Dortmund in the fourth and final quarterfinal matchup.
March Madness
Any fan of March Madness enjoys how over the course of about 20 days, there are 63 single-elimination games. I’ve found myself watching a bunch this year, and the drama has been compelling in a wild tournament full of upsets.
Omnifantasy drafters seemed to properly value the inherent randomness of this tournament structure, with only Gonzaga (19), Baylor (28.9), and Michigan (44.6) coming in as picks in the first seven rounds of Omni drafts that featured all 13 sports. As I write this, those three teams are all still alive, with Baylor having reached the Final Four to lock in 30 points and Gonzaga and Michigan both having reached 20 points and favored tonight to join Baylor.
After those top three teams, the uncertainty was apparently justified. Each of the next four highest-drafted teams — Villanova, Ohio State, Iowa, and Illinois — finished with zero Omni points. Houston was on average the eighth-highest drafted team, however, and if Gonzaga and Michigan do win tonight we’ll have a Final Four comprised solely of teams picked within the top eight, despite the many landmines. That would seem to follow a trend I mentioned in my post reviewing past Omni results, where each of the five overall NCAAB winners in my longest-running league have been top-eight NCAAB picks in our February Omnifantasy draft. Longshots can certainly make runs in this tournament, but the winner tends to come from a pool of the best teams that do survive the early gauntlet.
Of the other four point-scoring teams, Arkansas was the only one drafted reasonably high. The three Pac-12 schools were rarely picked, with USC being the only one selected in what we’ll call “standard” leagues for the below review of ADP. After Houston, the next seven highest-drafted schools all failed to score points before Arkansas came up, which is to say that 11 of the top 15 NCAAB picks in ADP recorded a zero, and then the other four very well might be the entirety of the semifinalists and beyond. This was a wild tournament with unprecedented upsets — my favorite stat was more 7-point dogs have won outright than any year since the field expanded — and it only served to reinforce that NCAAB is a boom-or-bust sport for Omnifantasy purposes.
Golf gets underway
We haven’t updated the golf results on the site yet because things are scored a little differently there, but I’d be remiss if I didn’t discuss the results of the first event. Justin Thomas took down The PLAYERS Championship in early March, and while that doesn’t guarantee him any Omnifantasy points, he does have a nice cushion to start the Omnifantasy golf calendar.
As a reminder, we score five tournaments — the four majors plus The PLAYERS — and an aggregated set of results in those tournaments determines who wins the 80, 50, and beyond for the larger Omnifantasy contest. After the first tournament, these are the results in “golf points,” which should be thought of as a contest within the larger contest. They don’t actually count for anything other than arranging the final golf results after we have all five tournaments completed.
We award “golf points” for the top 16 finishers in each tournament, and typically you’ll see fractional points awarded when ties dictate. As luck would have it, though, this first tournament had an outright runner-up, then breaks after fourth place, eighth place, and 16th place, so the points were awarded evenly.
It’s not a perfect system, but it does seem to work pretty well. What I can tell you from past results is that the eight golf points Thomas has already locked up have historically meant an overall finish in the top six, at worst. Thomas of course will have the ability to add to his eight golf points, but even if he doesn’t finish well in any of the later majors, he’s likely to have already secured 20 Omnifantasy points by winning one of the tournaments.
Westwood’s five points, meanwhile, have traditionally been right on the borderline of good enough for eighth place. Westwood wasn’t drafted in most Omnifantasy leagues, and if he does finish in the top eight, he’d in most cases “steal” Omnifantasy points from the field. Given where he’s at in his career, though, it wouldn’t be a shock if he went the rest of the Omnifantasy season without accruing any additional golf points, which very well might prevent him from finishing top eight and doing that.
I usually update the overall standings to reflect the current golf and tennis results once we have at least two tournaments in the bag.
Average Draft Positions and the optimal draft
Figuring out how to look at ADP is kind of tough in Omnifantasy since leagues are fully customizable. I’ve tried to encourage people to use all the sports they can, but it’s been very common for leagues to drop some sports. I’ve also seen plenty of smaller leagues, which are great, as well as some larger ones that rival even my longest-running league and its 16 participants. Those larger ones can’t use all the sports, since there are only 12 WNBA teams, and thus any league with more than 12 people can’t use WNBA.
To pull ADP data, we wanted to set certain parameters, so we went with leagues of between 8-12 people that used all 13 sports and did the standard five flex spots (18 draft rounds). That left us with 10 completed drafts, so the ADP data isn’t particularly robust, but there’s at least a decent sample. Here is a link to those results in a Google Sheet, which I’m unfortunately unable to embed.
The first thing that immediately stood out to me is that even with just 10 leagues, there were only two picks — Novak Djokovic and the Lakers — that went in the first round (top 12) of every draft. Djokovic and the Lakers were also the only picks that went in the 1.01 in these leagues, although I saw other picks go that high including Man City and the Chiefs.
Other picks that had relatively narrow ranges near the top included both Man City and Bayern Munich from the Champions League, both Alabama and Clemson in college football, and then the Chiefs and Dodgers. Outside these top eight options (including Djokovic and the Lakers), no other pick went inside the top 20 of every draft, and there were only four more picks (Seattle Storm, Brooklyn Nets, Rafael Nadal, and France in EURO) that were inside the top 30 picks of every draft.
The disparity between min and max pick for various options is fascinating given the sample is just 10 drafts. At the extreme, you have examples like Dustin Johnson, who went fourth overall in one league and as late as 42nd in another. The Milwaukee Bucks went at eight and 45. The Maple Leafs went third in a draft, and also as late as 65th. But my favorite of this genre is the Saints, who went as high as seventh overall in one draft, and were completely undrafted in two of the other nine leagues in this sample.
I wouldn’t say I expected anything different, as this is sort of the point of Omnifantasy. There are going to be some homer picks, but if you believe in something, you should go after it. Every draft is going to value the different sports in different ways and have runs and other extremes that create some chaos. You can be a value drafter if you want, but identifying the value is the hard part, or we’d all crush the futures markets all the time. Of course, when you do buy into a pick and it hits, it’s fantastic.
Kevin wrote a quick script using this ADP data and our expected points update from February 13 to determine an optimal 10-team draft. His script was just trying to maximize expected points, and he added this disclaimer:
This is not 100% guaranteed to be optimal... If it turns out that the optimal strategy involves taking, say, five picks whose ADPs are all in the seventh round, this strategy will miss this because, with each pair of picks, it's only looking at the picks whose ADPs fall between that pick and the pick two rounds later. However, ADP would need to be really far off for that to be the case, so I don't think we're missing that here.
So what happened? At Pick 1, the optimal option was Novak Djokovic. After that, with Djokovic off the table, the optimal first-round pick from every spot in the first round was the Seattle Storm, which makes some sense given the high expected points of the WNBA favorites in a league with only 12 teams. A question we could ask, of course, is whether WNBA futures odds are as reliable as other more heavily-bet markets.
Here’s the full draft from Pick 1:
And here’s from Pick 6, where early WNBA flexes became a major priority:
There are some interesting takeaways here:
Whether you buy the heavy WNBA lean, there’s probably something to the idea of using your flexes early, particularly if you identify a good opportunity in a particular tier of a particular sport. The first draft, which didn’t flex a WNBA team, still went back-to-back with the MLS in Rounds 5 and 6, and then back-to-back with NASCAR in Rounds 8 and 9.
Both drafts liked EURO picks in the first three rounds, and that seems like a clear sign given EURO was one of the only sports to push the optimal away from early WNBA picks.
Both drafts favored multiple MLS picks reasonably early, and it might be the case that my writing about how February MLS futures odds have historically not been very predictive started to impact ADP in a way that those picks appear to be values given the odds they do have assigned to them.
In both cases, the two sports the optimal punted until the last three rounds were PGA and NCAAF. The other sports the optimals didn’t prioritize, working backward from latest ADP selected, were NCAAB (126), MLB (111), NBA (102), and NHL (100). Of those, NBA is the one I’d say I was most surprised by.
Perhaps unsurprisingly, there were several picks that fit into both optimals — Chicago Sky (WNBA), Toronto FC (MLS), Ryan Blaney (NASCAR), New Orleans Saints (NFL), Houston Cougars (NCAAB), Barcelona (UCL), and Texas Longhorns (NCAAF). Of that group, Barcelona is the result of us pulling expected points before their first leg loss to PSG, which dramatically shifted their expected points and and correctly impacted their ADP in a way that is duping the optimal given the inputs we used (the ADP data is across the whole drafting period).
The rest appear to be value picks that for one reason or another drafters weren’t overly interested in. The Saints probably relates to Drew Brees’ expected retirement and what they will look like at QB next year, and I understood and continue to understand that apprehension, though their odds remain fairly solid. Texas (NCAAF) was apparently the best option for the optimal at a sport it ignored until the last round in both cases, while Houston (NCAAB) was probably undervalued as a smaller-conference team, and they’ve now locked in a Final Four berth to pay off those who went with them. In the earlier rounds, the Chicago Sky, Toronto FC, and Ryan Blaney all appear to have just been the best individual values in terms of expected points and ADP among sports the optimals wanted to target in the top-100 picks — all of WNBA, MLS, and NASCAR were taken in the double-digit picks in both drafts, and flexed there in at least one of them.
What’s on tap
You can always consult the “Schedule” tab in your league for a list of events, but we have an exciting early April on the Omnifantasy calendar including:
NCAAB — The Final Four on April 3 and April 5 as we declare the champion of our first sport.
Champions League — Quarterfinal first legs on April 6 and 7 and second legs on April 13 and 14, before the semifinals in late April and early May and the one-game final on May 29.
PGA — The Masters from April 8 until April 11 as the second of our five golf tournaments.
NHL and NBA — Regular seasons winding down before the playoffs start a little late this year, with the NHL first round beginning May 11 and NBA on May 22.
As always, results can be viewed in the individual sport tabs on the site, the Standings are updated, and the Big Board section has color-coded results by each drafters’ roster.
Love this write up and the omni format. In the Champions League they no longer redraw every round, instead drawing brackets for the semifinals at the same time they draw the quarterfinals. This means the winner of Man City-Dortmund will have to play the winner of Bayern-PSG, giving my early Man City pick a tougher road to the finals. Chelsea on the other hand gets to avoid all three top teams in the semis if they beat Porto, which further explains why their odds improved so much after the draw.
Love this shit Gretch