NBA Finals 2023:
In-Play Analysis


  • Usability Digital is aware of inconsistencies in availability of key in-play betting markets between sportsbooks
  • Usability Digital was unable to source any existing analysis in this area
  • Usability Digital also doubt any sportsbooks or data/odds suppliers have undertaken such analysis
  • Usability Digital experienced a high interest in this type of analysis following the release of UDEL’s Super Bowl LVII analysis, we would like to extend a similar analysis to other Sports


  • Conduct an in-depth analysis of multiple sportsbooks to establish in-play patterns and strategies
  • Conduct a sportsbook comparison of in-play user experience
  • Analyse the output


  • Usability Digital recognises that this information can challenge current in-play trading strategies and be used to enhance current in-play models
  • This analysis can also be applied to any sportsbook, betting market or pricing analysis model

In-Play Analysis: Basketball key in-play markets in the US

NBA Finals, June 7th, 2023: Nuggets @ Heat, Game 3

Purpose of analysis

  1. Compare the availability of three main NBA in-play markets across five major US sportsbook operators
  2. Gain insights into strategies and risk models utilized by major US sportsbooks for high-profile in-play wagering
  3. Monitor price change latency, availability for wagering, and price competitiveness of the moneyline, spread, and total across the selected sportsbooks
  4. The analysis focused on NBA Finals Game 3 between the Denver Nuggets and Miami Heat, which took place on June 7, 2023

In-Play Analysis: Assumption, Definition and Caveats


  • We assume that suspension incidents are the same for both web and app. We did spot checks on web/app and has yet to record any discrepancies between Web/App display of in-play markets
  • We used New Jersey sportsbooks and assume all sportsbooks in the comparison offer the same product and pricing in all states

Definition of in-play time

  • Market suspension has been defined as when a market is not available/visible on the front end. The market may well be ‘active’ in the backend and on a pre-populated betslip, but any player monitoring availability on the web or an app will not experience the market as available


  • All data was recorded and collected by Usability Digital Europe (UDEL) in good faith and without any preconceptions of what we would discover and without bias towards any operator or feed provider
  • All data is recorded, owned and stored by UDEL
  • We have taken every step to validate the information gathered
  • We believe the accuracy of the data is the truth as recorded but appreciate their may be other factors such as latency, rendering of data and information and other technical issues that may prevent all information to be available to UDEL
  • We appreciate there may be inconsistencies between data collected by Operators or Odds/Data Feed Providers and our findings. However, we will not always establish if these are related to issues at our end or at the Operator’s end.
  • We also welcome and appreciate any feedback from Operators and Odds/Data Feed Providers in case of errors or clarification. We will update the report accordingly if verified.
  • UDEL wholeheartedly apologise for any errors there may be and will be happy to rectify these UDEL will not accept any financial liability for the reports we have produced in good faith

Transition to in-play and unexplained market suspensions


  • There were no significant transition time to report in the NBA finals (in contrast to Super Bowl)
  • DraftKings experienced a lengthy transition time from pre-game to in-play (20:38:46 EST). Game total was unavailable for 49 seconds, while moneyline and spread were unavailable for 26 seconds
  • Shortly after tip-off on DraftKings’ web application, there was a brief anomaly where 1st quarter markets appeared instead of regular game lines (Popular Tab). This lasted for 5 seconds before the regular game lines reappeared
  • Two periods of multiple firms’ suspension occurred during the game. The first suspension, involving Caesars, MGM, and Bet365, happened at 21:04:11 EST. This likely occurred due to an erroneous suspension message or transition between quarters. However, the time taken by each book to restore betting availability varied: Caesars took 8 seconds, Bet365 took 27 seconds, and MGM took 12 seconds. The second incident took place at 21:39:07 EST, with the same three books suspending the main three markets again. Caesars suspended slightly after MGM and Bet365. This suspension also occurred between the second and third quarter

Availability of key in-play markets


  • Compared to NFL there were few occasions when betting on the main three markets were unavailable
  • Wagering opportunities only started to be withdrawn when Nuggets maintained an unassailable lead in the 4th quarter
  • Despite Denver Nuggets pulling ahead in the 3rd quarter all operators offered the main three markets throughout the period
  • FanDuel did not suspend betting during the first three periods. However, no Sportsbook suspended betting more than four times
  • Total suspension time was less than a minute for any sportsbook

Price Competitiveness (Snapshot)


  • We aimed to compare pricing strategies among six sportsbooks using two metrics:
  1. Over Round (a representation of a book’s theoretical margin)
  2. Competitiveness of the odds for the favourite (in this case Denver Nuggets, the pre-game favourite stayed odds on the whole game)
  • To analyze the pricing strategies, we opted to capture snapshots of the game every three minutes, recording the moneyline odds for both the Nuggets and the Heat

Headlines - Over Round

  • 30 cent lines were the norm for spreads and totals throughout the game at DraftKings, Caesars and MGM
  • At the game’s start, the moneyline overround varied among the Sportsbooks. FanDuel showed the least competitiveness (106%), while MGM and Unibet were also less competitive initially. However, Bet365 (104.6%) maintained a higher level of competitiveness with their pre-game odds.
  • FanDuel offered a 28c line for spreads and totals throughout the game

Headlines - Top Price Favorite

  • Sportsbooks may well have a pre-game strategy of staying with or against the favourite throughout in-play (moneyline and spread)
  • DraftKings was the most competitively priced sportsbook for a Nuggets Moneyline win, especially during the closely contested first three quarters of the game
  • Caesars were the most competitive price the Nuggets in the final stages of the game when it looked to be an inevitable Nuggets win
  • FanDuel was by comparison best price Miami Heat for most of the game
  • Bet365 and Unibet stood out by not offering the top price for either team throughout the game, indicating a potentially defensive trading strategy employed by both Sportsbooks
  • Unibet was the only sportsbook to offer whole numbers for spread and total
  • DraftKings moneyline odds were transposed for a brief period in the second quarter (less than a second). Nuggets went from -190 to +155 with the Heat odds reversing the other way. This occurred at 21:22:10 EST
  • The table below illustrates how many times each sportsbook had the best or joint-best price for the Denver Nuggets and the shortest price against them. Nuggets stayed favourite throughout the game and parity was only threatened for a very short period of time in the first quarter when Miami Heat were up 5 points
  • This may explain specific trading strategies employed for this game (Game 3 in the NBA Finals), but it is unlikely to represent an overarching approach to in-play betting
  • Unibet’s notable observation is their apparent strategy of consistently avoiding being the top price for any selection throughout the game. While there is no evidence to suggest this strategy was deliberate, it is likely aimed at minimizing arbitrage opportunities and deterring in-play bots

Price Move & Scoreboard Latency (Snapshot)


  • We wanted to explore which sportsbooks were quickest to change the odds after a significant play
  • There are approximately 99.3 possessions per team/per game in the NBA. Average points are close to 114 points. This yields less than 0.9 points per possession in the NBA, meaning a three-pointer being the most extreme outcome per possession (0,1,2 or 3 points can be scored). We then decided to measure odds changes after 3 pointers
  • Consequently, we decided to measure the average time it took for a Sportsbook to change its odds after a three-pointer. As we could not establish the real-time TV latency we decided to benchmark it against the sportsbook that visibly changed its odds the quickest following a three pointer (see example below following a Miami 3-pointer)
Bet 365 changes the odds first

Price Move Latency (Snapshot)


  • Bet365 and MGM were the quickest sportsbooks to update odds following a 3-pointer. They both averaged less than ½ second after the quickest change.
  • DraftKings (1.67 second delay) and Caesars followed with an average delay of around 1.5 seconds after the first book changed its price
  • FanDuel were significantly slower to update their prices on the front-end (they may have changed in the back-end) but it gives an impression they were well behind their competitors. The average FanDuel delay was over 5 seconds

Scoreboard Analysis


  • MGM was the only sportsbook that synchronized odds and scoreboard changes, consistently matching them perfectly.
  • Bet365 were also consistent with no difference between scoreboard and odds changes greater than 3 seconds Overall
  • Caesars’scoreboard was the quickest – always ahead of or matching the scoreboard speed of its rivals FanDuel had a significantly delayed Scoreboard update, taking an average of 12 seconds longer compared to Caesars
  • MGM’s clock management was confusing, appearing to lack a connection to an official data feed. The clock only advanced in 15-second increments and often required resetting when the official game clock was stopped
  • FanDuel’s Scoreboard encountered errors, as it displayed incorrect scores in the first quarter. The score changed back to 4-7 twice when the actual scores where 4-6 and 6-9.
Caesars always first to move the Scoreboard when scores changed

Timeout Analysis (Snapshot)


  • When reviewing gameplay we noticed some heightened activity during a timeout in the second quarter (5:36 left on the clock). Both Heat and Nuggets changed players (Duncan Robinson and Bruce Brown leave the court) after the timeout but this does not explain the many price changes made by some sportsbooks
  • During the timeout Caesars made 23 changes to the main three markets. DraftKings also made a high number of changes (18), whilst FanDuel did not make a single change
  • The frequent price changes, many within a couple of seconds of each other may impact on user experience as players may see ‘price change’ messages on their Betslip when they attempt to place a bet. Player betting with Sportsbooks with no bet delay and a ‘tolerance level’ (in case of odds change) for smaller wagers may not experience this inconvenience

Would you like to know more?

For more information about this analysis or to speak to our Sportsbook and Trading consultants, get in touch at