Thunder Q&A: Talking Lottery Odds with Tyler Carroll

If you’ve ever found yourself scrolling alongside Thunder Twitter, chances are you’ve happened upon Tyler Carroll.

Tyler, age 29, has transformed the Thunder Twitter fanbase from a herd of tank watchers to mathematicians through his sharing of data analyzing the franchise’s future come Tuesday’s lottery.

Tyler’s lottery findings have been a product of months worth of lottery simulations, tinkering, and reporting on all 13 possible pick combinations for the franchise as he’s concocted a 15,019 row, 5,500 column excel sheet (I am not joking) breaking down the likelihood of which of Mark Tatum’s envelopes will rest a Thunder logo.

After looking at Tyler’s excel sheet, I connected with him over email to discuss his work leading into Tuesday’s lottery. Here’s what he had to say:

Q: What drew you into figuring out the Thunder’s draft odds?

A: About halfway through the season when we knew the Thunder were tanking, the top 5 draft prospects had franchise-changing potential, and Houston had a pick swap possibility, I found myself going to Tankathon often and just clicking away seeing all the different outcomes. I always wondered the exact odds of us getting a top 5 pick, as I found out later A LOT of people on Twitter wondered the same too. I am an engineer and learned, and really enjoyed, coding in college. I thought it would be a really fun project for me to create my own version of Tankathon, but on steroids, so I could analyze the results.

Q: How do you run a simulation, and how many have you done so far?

A: In the excel file, you input: the Tankathon standings, win-loss records for each team to differentiate ties, and how many simulations you want to run from 1-15,000. Once you have these inputs, with a click of a button, the program I wrote spits out all the simulations to one of the sheets. I made formulas that search the outputs for Oklahoma City, Houston, and Miami, so it shows all of our pick swap possibilities. The only thing I really cared about in the beginning was how often do we get a top 5 pick, that later expanded into a lot of different things. But, I took the results from that and made it presentable on another sheet.

Once I had this, I ran 15,000 simulations for each OKC-HOU final standings combination, almost 500,000 simulations total, and put those results into a chart, and then into a nice graph. The graph pretty much tells the whole story… At this point, I knew I had something worth informing Thunder fans. We wanted Houston to be super-duper bad and us to be almost as bad to get the best odds at top 5 picks.

Once I had this graph, I would look at who played each night and figure out the different scenarios of each team winning or losing to see how it affected the Tankathon standings. I would take the best case and worst case from each night to see how much stake a top 5 pick was up for grabs each night. I would then update Oklahoma City’s top 5 pick percentages on Twitter after every game. I got a lot of great responses and suggestions on Twitter that helped me do more features and look at different things. Thunder Twitter is the absolute best!

Q: How did you come up with the combination odds?

A: During the season, I used the results from my simulations for the odds I reported, which are typically within +/- 1% of the actual odds, but we can say +/-2% to be safe. After the season, I actually calculated the odds using math with help from Alex Kelley.

Q: One of the big events leading up to Tuesday’s Draft Lottery was the May 25 coin flip that the Thunder won against the Cavaliers. How did that victory impact the draft statistics, if it all?

A: Winning the tiebreaker salvaged a 0.86% chance at two top 5 picks and a 4.78% chance at one top 5 pick… so a total of 5.64% at getting one top 5 pick [was gained]. It also helped on the backend, where the Thunder can only fall to 8 instead of 9 for the, almost, worst-case scenario. If Oklahoma City falls to 8, we are guaranteed Houston’s pick at 5. So, that makes the worst-case scenario pick 7 and pick 18 from Miami.

Q: What should Thunder fans be looking out for the most in terms of where the team will select?

A: The three most likely selections are picks 6 and 18 (18.07%), picks 5 and 7 (12.00%), and picks 5 and 6 (9.03%). The three least likely selections are picks 5 and 8 (2.4%), 4 and 18 (4.97%), and picks 3 and 18 (5.02%).

A full breakdown of all 13 pick combinations can be found here

Q: What are the chances Oklahoma City will nab Houston’s First Round Pick? Will certain Oklahoma City selections enhance the odds of Houston falling to 5?

A: Houston has a 47.9% chance to fall to 5. Yes, If OKC falls to 8, there is a 100% chance HOU falls to 5. If OKC falls to 7, there is a 66.67% HOU falls to 5. if OKC falls to 6, there is a 33.32% chance HOU falls to 5.

Q: A dream scenario for the Thunder would include them selecting at picks 1 and 5, is that occurrence likely on your data? What about any combination of one top 5 selection, and two?

A: Picks 1 and 5 occurred in 6.4% of lottery simulations. One top 5 pick showed 75.93% of the time, while two top 5 picks appeared in 24.93% of runs.

Q: You’re the most knowledgeable person on Twitter for OKC’s lottery odds, what’s your demeanor headed into lottery night? Are you going to be keeping tabs on odds as picks run by, or will you just be anxiously waiting as Mark Tatum announces selections?

A: Stevethunderfan may be more knowledgeable than I, but I appreciate the kind words! Stephen said that they will announce the lottery in reverse order from 14 to 1. So, I will be keeping tabs on the picks as they announce them. Count how many teams are skipped, the teams skipped jump into the top 4. We don’t really want to see any teams get skipped, the fewer skips the better for our odds at a top 4 pick. If three teams get skipped, maybe hope that another gets skipped so it guarantees us Houston’s No. 5 pick.

Q: Given you’ve done your due diligence on the Thunder’s draft odds, do you have any preferences as to who you’d like the team to select?

A: Without the intangibles, Mobley seems like the best fit — he’s a great pick and roll or pop partner with SGA so he’d be fun to see. Factoring in intangibles, I would say my big board would be from one to five: Cade Cunningham, Jalen Suggs, Jalen Green,  Evan Mobley, and Jonathan Kuminga.

If we get 2 top 5 picks, the dream would be Cade and Mobley. [Though it] most likely would be Kuminga and one of the others depending on where the higher pick lands.

Q: Anything else you’d like to add?

A: Special thanks to all the Thunder podcasts, especially the Down To Dunk guys that helped spread the word! Hope and pray for 1 and 5 and #ThunderUp!!!