If you roll stop signs you are my enemy.
Mainly post about Best Ball. Love a good graph. Triggered by red/green color schemes.
W&M + Duke. Data/BI Analyst
Finished listening to
@PatKerrane
's new Legendary Upside pod yesterday and wanted to touch on a topic I think got cruised over pretty quickly.
How important are bye weeks and how close is third place to second place typically?
🚨 I'll be writing best ball articles for
@4for4football
this offseason!
First up is an intro to best ball. Next, I'll be diving deeper into BBM strategy. Keep an eye out for that soon
I appreciate anyone checking this out or anything else I put out in the coming weeks
We are excited to announce that
@FeilNarley
is joining our team! 🚨
His first article is an INTRO to best ball leagues!
He broke down:
❓ Why you should play
🏗️ Structure
📖 Rules
📊 Where to play (UD/DK/Drafters)
👀 Strategy
Check it out:
I've certainly been an evangelist of Underdog, but I wouldn't really say the transparency is top-tier. I doubt there are major issues, but verifying that is difficult for several reasons.
Even the advancing Gauntlet data posted last week is unusable. Neither dataset is complete.
It’s undeniable at this point that
@UnderdogFantasy
is emerging as the industry leader in DFS, Pick’em, Best Ball, etc. Their product speaks for itself. Their team actually responds to customers. They operate with transparency. Their rakeback program is amazing compared to DK.
Looking through BBM data to check outliers often leads to seeing intriguingly uncommon teams.
Like this
@DavisMattek
gem of a draft
I like to think this was his first pick of the slow drafts on his honeymoon. Certainly living more than a little here
Some strong immediate corrections for several players. Any significant mover will likely continue to move in that direction at least a bit more. When ADP shifts there is always a slight lag
@SamHoppen
I’m very confused why so many people expected a non-unanimous first team all pro QB would be the unanimous MVP. Doesn’t make any sense to me that he got more MVP votes
I ran some pretty rough numbers on the impact of week 16 vs week 17. If you're able to move your win rate by the same factor, I lean toward week 16 being more important.
The most important and uncertain part to me is how much can we reliably influence each of these figures
I might post about how finals structure might change your incentives soon. My rough math and process for what I have in mind are far from perfect though. They serve more as very rough estimates
Looking at 2022 BBM data, it is beyond absurd how linear the relationship is between advance rate and draft capital spent at running back and wide receiver.
Doubt we ever see such linear and decisively positively or negatively sloped curves again
Getting deeper into NFL best ball after not doing much the last few months.
I was curious how much the first few days were filled with grinders. Surprisingly, a good amount of lower-volume drafters entered immediately. Also, the proportions are relatively steady until August
You don't think you have to know all the ins and outs of each player to advance above expectations in best ball tournaments.
All it takes is setting aside your ego and listening to the market.
How likely are advancing teams really to advance?
First place teams are quite likely to advance at this point but how far they are from the advance line also matters.
I'm only going to touch on ADP value mattering here because I've written about it before.
Largely, you do not have to predict the future. There is a relationship between real-time adp value and CLV. The correlation is ~0.64, not perfect but rather strong.
How much does average draft position (ADP) value impact playoff expectations? (ADP Value = Player pick - adp at the time the team was drafted)
After charting all 480K teams, we find that there is no advantage of ADP value
The best teams are those centered around 0 adp value
Interesting to see this update. Draft capital seemed a bit too flat to me before and this should be an improvement.
As a side note, as someone hesitant to reach out to others, I'll always appreciate how receptive and helpful Michael has been to the questions I send his way.
If anyone is interested, I made a new version of my Draft Capital Metric that weights the top picks a bit more heavily. Thanks to
@FeilNarley
for the help there / recognizing the steeper curve was more accurate.
Manifesto analysis still uses the initial version, but I added the
I'm just going to reiterate how much I think people are overstating the importance of ownership for the first round of the best ball playoffs.
Calling Mostert "chalk" seems a bit aggressive to me when he's more likely to be on three teams or less in your pod than four or more
Seeing who actually made the BBM IV finals now, I thought it would be interesting to see how the much talked about pod luck played a role.
I calculated the odds of each advancing team making the finals and compared that with the actual outcome.
It can be neat seeing some draft strategies come together.
-Henry
-Tez Walker @ 60?
-Overreacting to Chubb lifting @ 61??
-Jeudy/Watson @ 84/85???
What is this person doing? They’re utilizing the knowledge the learned in the depth of YouTube at Live a Little Fantasy Football
People seem relatively unsure, but optimistic the breakeven rate increases with the new structure.
From my math, it definitely does!
My calcs look at maxing the tournament. The median outcome decreases from ~1475 to ~1050, but the odds of coming out in the black rises
People who plan to max BBM, how do you think your odds of breaking even changed with the new payout structure?
To be clear, if your breakeven chances were 20% previously, a 3% increase would make them 23%
After talking with
@Sherman_FFB
about his analysis of flex utilization recently, I joined some weekly data from
@nflfastR
to look info how
@UnderdogFantasy
flex utilization actually broke down each of the last three years...
ICYMI - I wrote up some common trends between best ball tournaments for the last four years and thoughts on applying them this year.
More coming soon...
Stacking is paramount to having a chance and so is differentiation. This shows how frequently players are stacked with their quarterback when the QB is drafted.
I'd much rather have Rice with Mahomes, especially given Kelce is stacked much more frequently.
The field is unquestionably worse now than the previous weeks. I'll post a couple interesting tid bits before the playoff start and will track data as we advance. This is my final draft landscape thread though
Please, follow/like/share if you found this interesting/useful
Jeff Wilson Jr and Raheem Mostert are two of my top eight exposures. Regardless of how it plays out, I’ll always have
A) the thrill not being coffined today
B) what I expect to be some sweet sweet CLV
Look forward to losing this year but telling myself I did it the right way!
Ran this again but for actually winning BBM. The most likely team to win taking into account player level results took it down this time!
The next most likely teams to win bricked out earlier. Hope you don't look at this and let it live in your head indefinitely
Seeing who actually made the BBM IV finals now, I thought it would be interesting to see how the much talked about pod luck played a role.
I calculated the odds of each advancing team making the finals and compared that with the actual outcome.
On a run and don’t know if there’s materially new information but I don’t care about the risk this can’t be efficient.
This is a weekly winners draft fyi
The number of people who defend these approaches will always surprise me. You're limiting your lineup optionality and likely sacrificing more points than just the bye week.
Even after 14 weeks of accumulating points, the difference between advancing and not is often small
Who are the people that think solo QB or solo TE is a good strategy?
It's def not
It hurts your reg szn adv rate and kills any chance of uniqueness for the playoffs (i.e., if Kelce has a big reg szn to help you advance and big playoff weeks to get to w17, he's highly owned)
How do your teams compare to other
@UnderdogFantasy
Best Ball teams?
After week one this is how the point distribution in my leagues breaks down. I would not react too strongly in either direction if you're doing well or poorly because there's a lot of football left.
Collected data on my Mitten pods.
There's a TON of variance between pods (especially Mitten 1/2 with the pods of 6), but certain general trends emerge...
Yes, player performance also factors into luck, but I looked at how the pod formation.
Naturally, most people making the finals are lucky. My biggest surprise was that
@JustinHerzig
was actually the 10th unluckiest person, despite making it. He was 0.785 teams below expectations
@ChessLiam
I always think this is quite flawed thinking because it’s one player out of 18. Having a different 12th round pick doesn’t sound preferable. Mostert has been the best pick where he went and allows you to “drag” under performing players in that could do well, like Christian Watson
Want to see how the greater powers that be conspired against you??
Running the last couple of weeks with a few different pod formations, this is my approximation for the cut line distribution across pods for BBM this week.
Just as every post highlighted people advancing out of the regular season above expectation, the same will be true from this round to the semi-finals.
Advancing a team or a few across contests is often above expectation though. Consider BBM...
Took me a bit longer to get together than planned but covered a breadth of topics for best ball tournaments here
An extensive look at the historical data we have for BBM. Much of this should apply to all tournaments.
I will come out with more soon so stay tuned!
Are you ready to dive DEEP into Best Ball Mania?
ADP value, roster construction, position-specific approaches, rookies, and more!
@FeilNarley
provides the ultimate Underdog BBMV Draft Guide:
These were the most common pairings last year. Quite notable how much higher many combinations are this season.
Nice to see the data was released, it'll be interesting to see how else play has updated
Want to know if you actually have the right to complain?You can check the rest of the results in this google sheet,
Thought this was interesting so a bit behind on playoff best ball. I'll post later this afternoon about my drafts
If you're looking to draft any Mittens. Here's my last thread about the landscape.
Also, it's important to recognize that the advance structure is significantly different here with a 1/6 in the first round
After taking a month to thoroughly evaluate the "elite" tight end debates, I wrote up my thoughts on the top tight ends, historical performance of tight end builds and tight end spike weeks.
I know many have waited for my thoughts with bated breath
Check out the interlude between drafts. Pete graciously had me on to discuss the importance of ADP value and let me give a voice to the greater Bye Week Bro agenda!
Big win for the spreadsheet virgins!
As always continuing to look at some other stuff. Also, may replicate some of this with BBM and/or BBM II data.
If you found this interesting please give me a like, follow and/or retweet. Appreciate it and thanks for reading.
I understand why things like this are rolled out, but I'm not a huge fan.
There will be confusion. Some people (including people from UD) don't seem aware that the "vulture" has to be a non-RB.
TD prices are often extremely bad, esp on UD.
This is frequently a one-way market.
Coming next Wednesday… Vulture Protection 🙌
On Underdog, there will be No Vultures in Week 1 ❌
If you select a Designated RB to score a TD in your Pick’em Entry and they get vultured, we will Rescue your entry 🤝
Ran my first half in ages today and my friend ran his first.
It’s been fun getting back into running this year and pushing myself. From running 0 mile months to now a lot has changed. Hoping to keep up the miles and continue improving
If you thought this thread was interesting/helpful, follow me and like + retweet this.
I'll be posting similar breakdowns in the coming weeks.
I've also posted about BBM pods/advancement, if anyone is interested, check out my profile.
Thank you for taking the time to look through this. Please, like, share and follow me.
I'll be posting threads looking at the draft environment weekly moving forward. If anyone has questions they want to check out, please let me know.
Well, we all know who advanced and posted how great we've done so far.
Here is the distribution of advance rate for users with over 100 teams. Turns out the median advance rate was slightly higher than expected but somehow not everyone annihilated???
Completely burned this Achane/Waddle team, but if I ever opt for Baker/Carr over Achane/Waddle in a 2v2, it's over for me. I'm phoning it in. Withdraw my funds and buy me a casket
Sam brings up a good point here.
I’m looking forward to next week’s segment about Nico Collins going in the second round despite being out targeted and scoring less touchdowns than a 5’10’’ 165 lbs third round 24-year-old rookie from Houston in games they both played >50% snaps
They're very responsive to asks, but DK having csv downloads is objectively better for DFS. There are also more areas for exploitation based on the game setup.
With a proper backend, most exploits shouldn't be possible, but let's not mistake the transparency for being the best.
One annoying thing about starting to use BBM data is bringing in other data sources. This should help people start.
Joining data from
@nfl_data_py
or
@nflfastR
should be a smidge easier. If anyone notices errors from when I was collating this, lmk.
I'll post more about BBM results soon, but with the new data its interesting to check out the head-to-head battles.
Without knowing which drafts were streamed, we shouldn't read too much into this, but Justin Herzig looks positioned well against Dan Zack. Long way to go though
Posting a few tidbits about my playoff contests.
More heavily stacked teams are easier to get through earlier because more teams draft dead round-one players. Over 20% of teams from the earliest week cannot field a full lineup.
unavailable = byes + missed playoffs + AJ Brown
Researched how stacking and playoff game stacks influence advance rate using each BBM
I also took an extended look at the impact of byes on team performance with a different approach. I think this is more accurate regarding how much they influence advance rate
Best Ball: Leveraging the Schedule 📅
When a couple of points can make a major difference in the outcome, grinding the extreme edges can provide meaningful benefits
@FeilNarley
digs into EVERYTHING from stacking to positional weekly scores 👇
🧤🧤🧤🧤MITTEN 4 IS LIVE🧤🧤🧤🧤
Advancement Structure
2/6 Advance to Round 2
2/12 Advance to Round 3
2/20 Advance to Round 4 (314 team final)
You will also receive a Pick'em special this weekend for entering the Mitten 4
It's interesting how many people are confused by the possibility of an Aiyuk trade involving Kenrick Bourne.
It's simply a classic top-five wide receiver for a top-five wide receiver trade. Nearly identical receiver scores over the last three years
Dove deep into how the NFL schedule plays a part in constructing playoff teams. Some of my conclusions may be a bit different than you'd expect 👀
Come thinking I'll focus on game stacks. Stay for the essay on bye weeks
If we alter our view from the top of the queue to the top of the positional queue, we see a stronger more consistent trend emerge.
The value of taking from the top of the position queue outweighs the auto-drafting and poor structure in the previous chart.
@JustinHerzig
This is based on initial entries but you can get the teams advancing from the regular season by dividing n by 6.
If you had 25 teams in this round, >20% of the time you get 0 teams through. This is before you account for team-level correlation
Just as every post highlighted people advancing out of the regular season above expectation, the same will be true from this round to the semi-finals.
Advancing a team or a few across contests is often above expectation though. Consider BBM...
Being cognizant of ADP substantially impacts advance rate. Trying to differentiate in large field tournaments can make sense, but consistently eschewing ADP is not advisable.
Going full
@Dan__Zack
and "not having player takes" would likely behoove most players, myself included.
Slows only seems like an odd decision given that nearly every draft will have the clock decrease to four hours.
Be aware of that when you're entering. Clock changes next Thursday
Looking at advance rate data is very interesting in Best Ball Playoff contests
- Bye teams proved difficult to advance
- Hockenson did not advance much because of his stacking partners
- Josh Allen's advance rate varies significantly according to by week expectation at the time
The most important thing is making sure we're able to even field a finals team. Depending on your target final, the ideal number of available final players shifts because of byes but we need to always have this front of mind.
WAY less teams are viable for the finals now
FWIW, it can be frustrating, but this is probably the most entertaining format IMO. Could variance be reduced with wildcards?
Sure, and I might prefer it slightly, but people will always complain about what could have been. And that's fine
Here you can see which teams are commonly stacked together. ADP pushes players into similar teams pretty frequently.
Some stacks may require the room to cooperate, but taking some less common ones reduces the number of teams you're competing against