Experiment - What if RR14 Was Decided by Players' Personal Best Times?

A tournament where speed beats consistency

Posts with the Experiment tag are various playful projects, presenting different scenarios, new rules, and things that could have played out differently. Don’t take things too seriously in here, but do enjoy the alternate realities emerging!

Roulette Rivals requires competitors to play their best, round after round, with their current form being the difference between a deep tournament run and early elimination. But what if our players were machines, with no bad days whatsoever? What if players were compared by their best completion time on each map? Follow me into this alternate universe as we celebrate the RRStats team adding spin times for every known match¹ and the new Personal Best feature appearing on shoutcasts.

The rules for this experiment are simple. Every map and match of Roulette Rivals 14 is recalculated according to the players’ post-RR14 personal best times.² Thanks to CurryMaker and HITMAPS, I have all unused random maps for an authentic simulation.

Obviously, this simulation will result in players winning matches they originally didn’t. If two players face off in this bracket without meeting in RR14 first, they will be given six new random maps, underlined. This is a great help to players who had to forfeit the real tournament, now everyone has a fair shot at the title.

Round 1

51 players began Roulette Rivals 14, with 38 starting in Round 1 and 13 waiting in Round 2. Not many surprises are found early on, but three results are already different, those winners marked in purple instead of green.

Jokerj was one of the three dropouts in RR14 but his run is saved here, he sweeps through Ashton00122. Here CurryMaker keeps his foot on the pedal and wins two more maps after Mendoza, while Peter Dutton MP wins a faster and more exciting five-mapper with a clean Chongqing.

The Rieper 47 vs OhShitMan is the first example of the fact that map ties can still happen, they have identical personal bests in Paris, so the points are shared between the two.

Only three changes so far but the butterfly effect is only getting started…

Round 2

In our experiment, four matches play out differently in the first complete round of RR14. Jokerj and CurryMaker continue their WB run with 6-0 wins for both, while Peter Dutton MP faces off against Scruffy and loses.

In the toughest Round 2 match, Yannini’s experience prevails and advances over Pigiero in this experiment. Only one of this round’s matches goes past four maps, where eventually SkyL3R triumphs in a nail-biting 7-5 match against Veggerby. Apart from these changes, early rounds of the Winners Bracket feature most of the favorites advancing with ease.

LB Round 1 and LB Round 2

Throughout this experiment, the Lower Bracket will keep seeing more deviation than the Winners Bracket. Ashton00122 begins a successful LB run after being knocked down by Jokerj, while PurpleKey wins the battle of two dropouts.

Instead of LB2 elimination, several veterans find a way through the first obstacle: T_Nort23, OhShitMan, Rocky, and Some Random Person live to fight another day, while Ashton and Pigiero’s runs are just getting started. The match of the round has to be Rocky defeating Parapluie, the two players take eight maps in total³ before arriving at a score of 7 to 5.

LB Round 3

Before catching up with the Winners Bracket players, another elimination round has to take place, which features an OhShitMan win over T_Nort23, BigMachete overcoming Vulcat after two tied maps, and mikulers defeating Ashton00122 in five. The RR champion’s smooth LB run continues, which was Yannini’s role in our reality, but Pigiero’s doing just as well in this experiment.

Round 3

The final sixteen of the Winners Bracket meet in Round 3, with one notable upset. Jokerj, who didn’t even participate in the tournament, finds a way against In4Fun to defeat him 6 to 4. Jokerj has a better PB on only 8/19 maps but finds the right RNG to lengthen his WB run.

Another match that goes to five maps is Lord Munk vs lukedotpng, with the former eventually winning the back-and-forth. Almost all other matches are 6-0 sweeps this round, as while Phanium vs Dein Nomos was a tense five-mapper in our reality, the American’s personal bests are just too strong for the consistent Dein here.

LB Round 4 and LB Round 5

Though Music Inc only had a win on 10 of the 19 maps, he has the perfect map pool for his LB4 match against mikulers to defeat him. Every other match from LB Round 4 is a sweep, although quatilyti and lukedotpng’s matches could have easily gone the other way as they only had the better PB on 11 maps.

Thankfully, LB Round 5 is more exciting, and by this point, readers can notice that barely any matches are won by the corresponding RR14 players. ThatObserver is a sizeable favorite against quatilyti here, but he almost doesn’t manage to close out the match, not before an exact Whittleton Creek tie between them.

Round 4

Jokerj’s impressive Winners Bracket run ends in the Quarterfinals, losing to Phanium, who lost his real Round 4 match to In4Fun. Lord Munk pulls off the biggest upset of the tournament against Moo: he only has a better PB on four maps, but three of those four are a part of his winning map pool.

The other two Semi-Finalists are The Rieper 47 and Scruffy, winning their Round 4 matches against Frote7 and Yannini respectively, only losing one map each.

LB Round 6 and LB Round 7

By this point, our experiment is so far from reality that five of the next six Losers Bracket matches must be redrawn with random maps. Moo and Yannini pull off an equally sized upset, winning three maps despite only having a winning PB on 7/19.

In the other half of the bracket, Jokerj bounces back after losing to Phanium, and Frote7 ends lukedotpng’s run with a sweep.

The four winners then face off for the final two places in the final six, with Yannini converting his PB advantage into a win, and Frote7 scoring an upset over Jokerj. Joker’s run would end up being the biggest overperformance compared to his actual RR14 run, but it’s important to remember he forfeited the real tournament.

The Final Six

And so we reach the final six, with all randomly generated maps, except Phanium vs The Rieper 47, who met in LB Round 9 in the real tournament. Phanium and Scruffy both easily win their Round 5 matches, before tournament-defining encounters between them.

The Rieper 47 and Yannini have an identical record against each other, this time the three-time champion is the one advancing, he would face Frote7 after an easy win over Lord Munk.

It’s not a real experiment without a Yannini vs Frote7 matchup and sure enough, they find each other in LB Round 9. Frote7 has the narrowest of leads but Yannini’s maps are favored more, earning the German a spot in the LB Final.

According to the public spreadsheet, the two players with the strongest Personal Best records were Scruffy and Phanium, well ahead of In4Fun, Yannini, and Frote7⁴. There are no unfortunate upsets for either of the favorites, so they inevitably meet in the WB Final, with Scruffy winning it 9-5, though not before a strong Phanium semi-comeback.

Phanium then has to earn his way back into the Grand Final, which he easily secures with an 8-0 win over Yannini. It didn’t matter which champion he had to meet, he would have had an insane 15-4 map advantage over either of them.

The experiment’s 100th and final match is no doubt the most exciting. Scruffy gets off to a 6-0 lead before Phanium equalizes. This happens again, 8-6 for Scruffy goes into 8-8 and a final map. The championship is decided in Whittleton Creek, with Phanium winning it by just nine seconds, taking his first (theoretical) title since RR11. And the final catch? Scruffy had the faster PB on 11 of the 19 maps…

Summary

While Roulette Rivals 14 was won by Scruffy, ahead of The Rieper 47 and In4Fun, this experiment ended with a Phanium title over Scruffy and Yannini. 30 of the 51 players finished the alternative tournament in a different position than in our reality, with Jokerj and Yannini pulling off the biggest improvements, surviving six and five more rounds respectively.

Long-time champions Phanium and Frote7 also got a nice deal out of this experiment, while players like ThatObserver, Some Random Person, and OhShitMan weren’t hindered by an off-day and cruised further in the brackets here.

On the other hand, one-off results made even more players exit the tournament earlier here than in reality. With Jokerj sending BigMachete13 down to LB, Vulcat found and lost to a tougher opponent early, and Veggerby saw the same fate against OhShitMan. But the key takeaway from here is that PBs aren’t as strong for the more consistent players, with The Rieper 47, Dein Nomos, and In4Fun becoming the three players who exited the earliest compared to their actual RR14 runs.

While Personal Bests aren’t everything, they’re good indicators of a player’s performance on an upcoming map or a match. And with these PBs appearing on the shoutcasts, spectators can now be more informed about matches than ever.

How did you find this experiment? What would you like to read about next? Let me know in the comments! Hitman Substacks will return in September, at the end of RR15.

1. With invaluable help from Scruffy, ThatObserver, Phanium, and of course RRStats creator CurryMaker.

2. For full precision, a pre-RR14 database should be used but one wasn’t available, and this way the RR14 rookies are better represented anyway.

3. Maps 6 and 7 were Hokkaido and Miami respectively, but neither of them had a PB at the time of writing this.

4. TheTimeCube is fourth on this list, ahead of Yannini, but he didn’t participate in RR14.