* With a minimum 12 games played. Share is out of possessions the team had while that player was on the court.Possession share includes possessions used via field goal attempts, drawing fouls, turning the ball over, passing for assists and extending possessions with offensive rebounds.Source: Basketball-Reference.com 2017Cavaliers1691Warriors18509.6 Unfortunately for LeBron James and the Cavs, though, none of the underdogs on the list above ended up winning the championship. That’s because the NBA Finals are particularly unkind to underdogs. We can split hairs about how much of a favorite Golden State should be, but no matter how you slice it, upsets of this magnitude basically never happen on this stage.LeBron James’s share of team possessionsJames has built an entire career out of doing everything for his teams: scoring, distributing, rebounding, defending and countless other little on-court acts that help you win games. But in these playoffs, his workload is approaching a level that’s unprecedented even by his standards.Through a combination of shooting, ball handling and rebounding, LeBron has personally been responsible for about 38 percent of the Cavs’ possessions when he’s on the floor in these playoffs. The only player (minimum 12 games played) who’s handled a higher percentage of possessions in any postseason since the merger? James himself in his 2015 playoff campaign, when he nearly willed an undermanned Cavs squad past the Warriors: 200176ers1592Lakers17687.8% 1993Michael JordanCHI38.029.45.036.8 2003Nets1624Spurs174615.1 2013Carmelo AnthonyNYK38.09.32.535.7 2018Cavaliers1611Warriors171019.6 2016Russell WestbrookOKC34.352.46.936.7 1999Knicks1631Spurs174516.5 1992Michael JordanCHI184.108.40.2065.7 The biggest NBA Finals underdogs since 1977According to probabilities generated by pre-series Elo ratings 2014Russell WestbrookOKC34.340.57.335.6 2018James HardenHOU36.735.92.537.2 UnderdogFavorite 1996SuperSonics1695Bulls183212.8 Source: Basketball-Reference.com Last summer, we speculated that the NBA was getting more interesting, if not more competitive. That premise ended up mostly holding true this season: Although it’s Cleveland and Golden State in the NBA Finals once again, their fourth consecutive rendezvous was also nearly called off. The Houston Rockets and Boston Celtics helped make the Warriors and Cavs work harder to get here than they ever had to before.1The two teams played 35 combined games to get through the conference playoffs, the most they’ve had to play in any of their preceding Finals runs. Another rematch looked far from inevitable just a handful of days ago — that part was interesting! (As long as you put aside the generally lopsided nature of most games this postseason.)Now we’re left with the matchup that has become as much a part of late spring as commencement speeches and pollen allergies. And although this year’s version contains many of the same characters as earlier sequels, there are just enough possibilities here to keep things, well, interesting — albeit probably still not competitive. Here are six numbers to keep an eye on as we see whether the Warriors can hang onto their title, or if the Cavs can shock the world again.The Vegas oddsIs this the biggest NBA Finals mismatch ever? According to the Las Vegas bookmakers, it’s in the conversation. The Westgate SuperBook installed the Warriors as 1-to-10 favorites going into the series, which translates to about an 89 percent probability of winning after adjusting for the vigorish. Using the archived numbers at SportsOddsHistory.com, which go back to the 1998-99 season, the only Finals matchup more lopsided than this one came in 2001, when the Los Angeles Lakers had an implied 94 percent probability of beating the Philadelphia 76ers. (The Lakers ended up cruising to victory in five games.)Our own Elo model is slightly more optimistic about the Cavs’ chances. Based on both teams’ pre-series ratings, Cleveland has roughly a 20 percent probability of beating the Warriors. (Our interactive model gives the Cavs a slightly better chance because it takes into account things that Elo alone ignores, including playoff experience and travel distance.) That’s still the eighth-lowest of any Finals underdog since the 1976 ABA-NBA merger, but it’s actually about double what the Cavs’ odds were heading into last year’s Finals — and only a bit worse than Cleveland’s 27 percent probability before the 2016 Finals (which they won, of course, in one of the greatest upsets in NBA history). LeBron’s workload is historic (again)Largest share of team possessions an individual player was responsible for in the playoffs,* 1977-2018 2003Allen IversonPHI36.736.02.536.2 2002Nets1601Lakers171716.2 2009LeBron JamesCLE36.439.54.337.0 2014Heat1638Spurs173020.9 1981Rockets1573Celtics166820.4 2018LeBron JamesCLE35.946.73.637.6 SeasonTeamElo RatingTeamElo RatingUnderdog Win % 2015LeBron JamesCLE37.645.24.9%38.8% SeasonPlayerTeamUsage %Assist %Off. Reb. %Poss. % 1986Rockets1640Celtics18078.7 The bad news for James is that his supporting cast this season is even worse than it was that year (or in any of his other NBA Finals seasons, 2007 included). So he’ll need to keep shouldering this historic workload through the Finals if the Cavs are to have any shot at winning. James is a superhuman athlete, but between his 41.3 minutes per game, his 38 percent possession usage on offense and his likely defensive responsibilities — according to Second Spectrum, no Cavalier defended Kevin Durant for more possessions in last year’s Finals than James did — it’s fair to wonder how much more of this The King can handle before running out of gas.The Warriors’ assist-to-turnover ratioMy colleague Chris Herring once described the Warriors’ offense as “beautiful chaos,” a system of intricate off-ball screens designed to spring their many skilled shooters free for open shots. In order to work properly, though, that offensive machine requires a lot of patient and precise ball movement, which Golden State has had a bad habit of getting away from at times this year. While the Warriors did lead the league in assist-to-turnover ratio during the regular season, they frequently fell victim to stagnant offensive motion and careless passing against the Rockets, who held Golden State to a ridiculously low 1.1 assist-to-turnover ratio in the three games Houston won during the Western Conference finals — a mark that would have easily been the NBA’s worst during the regular season.Of course, the Warriors had a sky-high 2.1 assist-to-turnover ratio in the games they won over Houston, a sign of how well their offense still functions when it really clicks. But replicating that will also mean cutting out another of the bad habits Golden State slid into against the Rockets: too much iso-ball with Durant. Nobody runs more isolation plays than Houston, and somehow James Harden and friends convinced the normally free-flowing Warriors to do the same, with ugly results. After running only 11.0 isolations per 100 possessions during the regular season (according to Second Spectrum), the Warriors were up to an astounding 28.5 per 100 in the West final. Durant is a brilliant 1-on-1 player, and sometimes that type of offense is unavoidable, but the Warriors are at their best when these plays are selectively mixed in amid the beautiful chaos — not when they’re the centerpiece of the attack.The Cavaliers’ 3-point percentageAs I’ve written before, Cleveland is abnormally reliant on 3-point shooting to power its streaky offense — and to compensate for a defense that ranked next-to-last in efficiency during the regular season. In the postseason, the Cavs’ 3-point percentage has been 10.5 percentage points higher during wins than during losses (unsurprising from the team that had the league’s biggest regular-season gap). When the shots are falling, Cleveland can beat anybody. But it’s anybody’s guess as to whether that will be true on any given night.Nobody typifies this Cavs phenomenon more than Kyle Korver and JR Smith, a pair whose value is almost completely dependent on how well they shoot the basketball. In playoff wins this season, they’re shooting a combined 47 percent from downtown; in losses, that figure drops to 26 percent. This might be a chicken-and-egg thing: Do the Cavs win because Korver and Smith shoot better, or do Korver and Smith shoot better because the offense is working better overall? There could be something circular there. But it’s telling that the quality of looks the pair gets (as measured by Second Spectrum’s quantified shot quality) barely changes between wins and losses — rather, the difference is almost entirely driven by big fluctuations in shot-making after controlling for the difficulty of their shots.That makes the Cavaliers dangerous (and frustrating) for fans and haters alike. Although the Cavs’ hot-and-cold shooting touch might not matter as much against a team as talented as the Warriors — Cleveland got demolished in last year’s finals despite matching Golden State’s 3-point percentage — one of the Cavs’ best paths to victory rests in one of their patented hot streaks.Golden State’s third-quarter runsAs our ESPN colleague Baxter Holmes wrote earlier this month, one of the Warriors’ deadliest weapons is their ability to go on a devastating run in the blink of an eye that buries opponents before they even know what hit them. Although it can strike at any time, it often manifests itself right after the team emerges from the locker room for the second half: Golden State’s third-quarter scoring margin during the regular season was 199 points better than that of any other team in the league.2In other words, the difference between the Warriors’ third-quarter scoring margin and that of the next-best team would itself rank second in the league in third-quarter scoring margin! And in the playoffs, the Warriors have outscored opponents by 130 total points in third quarters, versus only 20 points in every other quarter combined. In Games 6 and 7 of the Western Conference finals, the Rockets watched their season slip away largely on the strength of massive runs staged by the Warriors during the third quarter.To a certain extent, there isn’t much that Cleveland — or anyone — can do to combat the Warriors’ quick-strike tendencies. But for a Cavs team prone to wildly up-and-down sequences of play (both from game to game and within the same contest), keeping Golden State from being able to capitalize on vulnerable moments will be a victory in itself. (For what it’s worth, the Cavs actually outscored the Warriors by 4 points in the third quarters of the 2016 finals.)Two deadly lineups?A decent chunk of the Warriors’ dominance over the past half-decade stems from the success of a few specific five-man units — matchup nightmares for whom opponents have no good answer. That has carried over into these playoffs, in which the so-called “Hamptons Five” lineup of Stephen Curry, Draymond Green, Andre Iguodala (who will miss Game 1 of the finals but could be available later on), Klay Thompson and Durant has outscored foes by 22.9 points per 100 possessions thus far. In concert with Golden State’s switch-heavy defensive scheme, the versatility and length of that group ensures that the Warriors don’t suffer defensively even while playing their top offensive players together.But the Cavaliers have a lineup that has somehow been even more effective in the playoffs than the Hamptons Five (although in fewer minutes): The group of James, Smith, Jeff Green, George Hill and Tristan Thompson, which is outscoring opponents by 25.7 points per 100 possessions during the postseason so far.Lineup data is so noisy that there’s no guarantee that a given group’s apparent synergy in the past will carry over into the future. But that’s all part of the bargain with this Cavaliers team. Since they remade their roster at midseason, they’ve been using the playoffs as a lineup laboratory of sorts, searching for the group that works best together — and it’s still a work in progress. None of Cleveland’s other common postseason lineups (among those that have played at least 50 minutes together) are in the same neighborhood as the group above, though, while Golden State has four separate combinations (including the Hamptons Five) that rank higher than the Cavs’ second-best unit of James, Hill, Smith, Korver and Kevin Love.When you’re as heavy an underdog as the Cavs are, experimentation might be the best option, so we’ll see what group(s) coach Tyronn Lue turns to as the series takes shape.Check out our latest NBA predictions.