Promotion Race
The Race for Two: Suwon Samsung, Suwon FC and Seoul E-Land, Scenario by Scenario
Two clubs go up automatically this season. Three are in the conversation: Suwon Samsung on 50 points, Seoul E-Land on 45, Suwon FC on 44 with a game in hand. Daegu on 43 are alive in theory and take one of the two places in...
Two clubs go up automatically this season. Three are in the conversation: Suwon Samsung on 50 points, Seoul E-Land on 45, Suwon FC on 44 with a game in hand. Daegu on 43 are alive in theory and take one of the two places in about 7 percent of simulations, but the race is essentially three-way, and the head-to-head part of it is already over except for one match: Suwon Samsung at Mokdong on Saturday. After that, nobody in the top three plays anybody else again. The race will be decided by who drops points against the other fourteen clubs.
This piece takes the same 20,000-run simulation behind Final Stretch Revisited and asks it the questions a Seoul supporter actually asks. What does Seoul’s total need to be? What has to happen to Suwon FC? Which of their matches should we be watching? How much does Saturday change? All probabilities are the model’s; all of them are conditional on the information available on September 7.
The short version
- The most likely pair to go up is the two Suwon clubs, at 56 percent. Seoul with Suwon Samsung is 25 percent. Seoul with Suwon FC, meaning Suwon Samsung collapse, is 11 percent.
- Seoul’s total is the whole story. At 60 points, Seoul finish in the top two 28 percent of the time. At 62, it is a coin flip. At 64 it is 82 percent. At 67 it is settled.
- Suwon FC’s expected finish is 62, and they finish on 61 or fewer 43 percent of the time. That is the number Seoul are chasing, not Suwon Samsung’s.
- The tiebreak is against Seoul. When Seoul and Suwon FC finish level on points, Suwon FC are above nine times in ten, because the first tiebreak is goals scored and they lead 46 to 43.
- Saturday is worth about 34 points of probability. Beat Suwon Samsung and Seoul’s top-two chance is 55 percent. Draw and it is 31. Lose and it is 21.
The three run-ins, match by match
The “predicted” column for each club is built the same way as the Seoul sheet in the previous article: the club’s expected record first, then each result placed where it is most likely. Expected goals are for the club first, opponent second.
Suwon Samsung, 50 points, eight to play (bye Round 27)
| Round | Weekend | Fixture | Win | Draw | Loss | Predicted | Expected goals | Expected points |
|---|---|---|---|---|---|---|---|---|
| 26 | Sep 12 | Seoul E-Land (A) | 30% | 32% | 38% | Loss | 1.1 - 1.2 | 1.23 |
| 28 | Oct 10 | Ansan (H) | 80% | 14% | 5% | Win | 2.6 - 0.6 | 2.56 |
| 29 | Oct 18 | Hwaseong (H) | 50% | 32% | 18% | Draw | 1.3 - 0.6 | 1.82 |
| 30 | Oct 25 | Seongnam (A) | 51% | 30% | 19% | Draw | 1.4 - 0.8 | 1.82 |
| 31 | Nov 1 | Daegu (A) | 39% | 33% | 28% | Draw | 1.2 - 0.9 | 1.50 |
| 32 | Nov 8 | Yongin (H) | 59% | 25% | 16% | Win | 1.8 - 0.8 | 2.02 |
| 33 | Nov 22 | Gyeongnam (H) | 61% | 27% | 13% | Win | 1.6 - 0.6 | 2.08 |
| 34 | Nov 29 | Jeonnam (A) | 56% | 27% | 17% | Win | 1.6 - 0.8 | 1.94 |
Predicted record 4-3-1 for 15 points, expected 15.0, projected finish 65. Their soft spots are the two away trips to top-half sides, Daegu and Seongnam, and the Hwaseong visit. Their title chance is 60 percent; their chance of missing the top two altogether is 14 percent, and almost all of that runs through losing at Mokdong.
Seoul E-Land, 45 points, eight to play (bye Round 30)
| Round | Weekend | Fixture | Win | Draw | Loss | Predicted | Expected goals | Expected points |
|---|---|---|---|---|---|---|---|---|
| 26 | Sep 12 | Suwon Samsung (H) | 38% | 32% | 31% | Draw | 1.2 - 1.1 | 1.45 |
| 27 | Sep 19 | Daegu (H) | 47% | 30% | 23% | Draw | 1.4 - 0.9 | 1.71 |
| 28 | Oct 9 | Gimpo (A) | 54% | 28% | 19% | Win | 1.6 - 0.8 | 1.89 |
| 29 | Oct 18 | Gimhae (H) | 83% | 13% | 4% | Win | 2.6 - 0.5 | 2.61 |
| 31 | Oct 31 | Yongin (A) | 43% | 28% | 28% | Loss | 1.4 - 1.1 | 1.59 |
| 32 | Nov 7 | Jeonnam (H) | 68% | 21% | 11% | Win | 2.1 - 0.8 | 2.25 |
| 33 | Nov 22 | Chungnam Asan (H) | 59% | 26% | 15% | Win | 1.7 - 0.7 | 2.03 |
| 34 | Nov 29 | Gyeongnam (A) | 43% | 32% | 25% | Draw | 1.2 - 0.9 | 1.61 |
Predicted 4-3-1 for 15 points, projected 60. Explained in full in the previous article.
Suwon FC, 44 points, nine to play (no bye left)
| Round | Weekend | Fixture | Win | Draw | Loss | Predicted | Expected goals | Expected points |
|---|---|---|---|---|---|---|---|---|
| 26 | Sep 13 | Cheonan (H) | 62% | 24% | 14% | Win | 1.9 - 0.8 | 2.10 |
| 27 | Sep 19 | Jeonnam (A) | 52% | 26% | 22% | Draw | 1.7 - 1.1 | 1.81 |
| 28 | Oct 9 | Hwaseong (H) | 48% | 30% | 22% | Draw | 1.3 - 0.8 | 1.75 |
| 29 | Oct 17 | Seongnam (A) | 46% | 29% | 26% | Loss | 1.5 - 1.1 | 1.66 |
| 30 | Oct 25 | Chungnam Asan (H) | 62% | 24% | 14% | Win | 1.9 - 0.8 | 2.09 |
| 31 | Nov 1 | Ansan (H) | 79% | 14% | 7% | Win | 2.8 - 0.7 | 2.52 |
| 32 | Nov 7 | Gyeongnam (H) | 58% | 26% | 16% | Win | 1.7 - 0.8 | 2.00 |
| 33 | Nov 21 | Gimpo (A) | 53% | 25% | 21% | Draw | 1.8 - 1.0 | 1.86 |
| 34 | Nov 29 | Chungbuk Cheongju (H) | 73% | 17% | 10% | Win | 2.5 - 0.8 | 2.36 |
Predicted 5-3-1 for 18 points, expected 18.1, projected 62. This is the run-in that decides Seoul’s season. Six of nine at home, where Suwon FC have not lost (6 wins, 4 draws), and every one of the nine is against a club currently seventh or lower except Hwaseong. The model gives them at least two points per game. The only three matches where a Suwon FC win is under 50 percent are Hwaseong at home (Cha Du-ri’s swamp football, 48 percent), Seongnam away (46) and Gimpo away (53, with Gimpo’s home record the reason it is that high).
Two things keep this from being a formality. Frizzo, their 13-goal top scorer, lost consciousness in a collision on September 5 and is doubtful for Sunday. And five of their regulars, including captain Han Chan-hee, build-up midfielder Gu Bon-cheol and both strikers, are on four yellow cards, one from a one-match ban, with nine matches to get through. The model prices in Frizzo’s absence for Round 26 only. Neither risk is in the numbers beyond that, and both are real.
Scenario one: what total does Seoul need?
This is the cleanest way to think about the race. Take Seoul’s final total and ask how often that total was enough for second place across 20,000 simulated seasons, given everything Suwon Samsung and Suwon FC did in the same seasons.
| Seoul final total | Points from last eight | Chance of top two |
|---|---|---|
| 57 | 12 | 4% |
| 58 | 13 | 9% |
| 59 | 14 | 17% |
| 60 | 15 | 28% |
| 61 | 16 | 39% |
| 62 | 17 | 56% |
| 63 | 18 | 69% |
| 64 | 19 | 82% |
| 65 | 20 | 91% |
| 66 | 21 | 95% |
| 67 | 22 | 98% |
Read it this way. The predicted sheet (60 points) is a third-place total more often than not. Sixty-two, which is 17 points from eight, or something like five wins, two draws and a loss, is the coin flip. Sixty-four, six wins and a draw, makes Seoul the clear favorite for second. Sixty-seven closes the door.
Seoul reach 62 or more in 35 percent of simulations, 64 or more in 17 percent. That gap between “what we need” and “what we usually get” is why the headline top-two number is 37 percent.
Scenario two: what has to happen to Suwon FC?
Suwon FC can reach a maximum of 71. The model expects 62. Here is how often they finish at or below each total, and what it means for Seoul.
| Suwon FC finish at or below | Chance | Points they must drop (of 27) | What Seoul need to be second |
|---|---|---|---|
| 57 | 10% | 14 or more | 58 (or 57 and win the tiebreak) |
| 59 | 23% | 12 or more | 60 |
| 60 | 32% | 11 or more | 61 |
| 61 | 43% | 10 or more | 62 |
| 62 | 53% | 9 or more | 63 |
| 63 | 64% | 8 or more | 64 |
And the same relationship from the other side:
| If Suwon FC finish on | Share of simulations | Seoul top two |
|---|---|---|
| 57 or fewer | 10% | 73% |
| 58 to 60 | 22% | 55% |
| 61 to 63 | 32% | 32% |
| 64 or more | 37% | 21% |
The practical translation: for Seoul’s predicted 60 points to be enough, Suwon FC need to drop 12 of their remaining 27, meaning four non-wins including at least one defeat, or three defeats. Their sheet above has four non-wins already (Jeonnam, Hwaseong, Seongnam, Gimpo), which is why the model has them on 62 rather than 65. One more slip than that and 60 is live. Two more and 60 is probably enough.
Where those slips are most likely, in order: Seongnam away (54 percent chance they fail to win), Hwaseong at home (52), Gimpo away (47), Jeonnam away (48). Hwaseong is the one to circle. It is Suwon FC’s only remaining match against a top-six side, on a Friday afternoon public holiday, and Hwaseong under Cha Du-ri concede 0.9 goals per away game while winning few and drawing many, which is exactly the result Seoul want.
The tiebreak works against Seoul. K League 2 separates level teams by goals scored, not goal difference. Suwon FC have 46 to Seoul’s 43, and they score 2.0 per game to Seoul’s 1.8. In the 8 percent of simulations where the two finish level, Suwon FC are above in 90 percent of them. In 5 percent of all seasons the level finish is the one that decides second place, and Seoul win it one time in ten. Treat “level with Suwon FC” as third.
Scenario three: Saturday
Round 26 is the only remaining match between two of the three, and the only match on the list where a Seoul win takes points directly off a rival. The model’s three branches:
| Saturday result | Chance | Seoul top two | Seoul champions | Suwon Samsung champions | Suwon FC top two |
|---|---|---|---|---|---|
| Seoul win | 38% | 55% | 22% | 45% | 63% |
| Draw | 31% | 31% | 8% | 61% | 72% |
| Suwon Samsung win | 31% | 21% | 2% | 78% | 74% |
A win on Saturday nearly doubles Seoul’s chance of going up automatically and puts the title itself into play at better than one in five. A defeat does not end the race, but it moves Seoul from chasing Suwon FC to needing them to collapse: 21 percent from there is roughly “Seoul get 63 or more and Suwon FC still stumble.” It is also, note, the only result in the table that makes Suwon FC’s job easier, because Suwon Samsung pulling clear of everyone takes one of the two places off the board.
Suwon Samsung’s own view of Saturday: it is the match most likely to cost them the title. Lose it and their championship probability drops from 60 to 45 percent; win it and it climbs to 78.
Scenario four: the matches to watch that Seoul are not playing
For every remaining fixture involving Suwon Samsung or Suwon FC, the model asked: how does Seoul’s top-two chance change depending on whether the rival wins? The gap between the two answers is how much that match matters to Seoul.
| Round | Weekend | Fixture | Rival win chance | Seoul top two if rival wins | Seoul top two if rival does not | Swing |
|---|---|---|---|---|---|---|
| 26 | Sep 12 | Seoul v Suwon Samsung | 31% | 21% | 44% | 23 pts |
| 34 | Nov 29 | Suwon FC v Chungbuk Cheongju | 73% | 34% | 46% | 13 |
| 26 | Sep 13 | Suwon FC v Cheonan | 62% | 32% | 45% | 13 |
| 27 | Sep 19 | Jeonnam v Suwon FC | 51% | 31% | 43% | 12 |
| 31 | Nov 1 | Suwon FC v Ansan | 78% | 35% | 46% | 12 |
| 32 | Nov 7 | Suwon FC v Gyeongnam | 58% | 32% | 44% | 11 |
| 28 | Oct 9 | Suwon FC v Hwaseong | 48% | 31% | 42% | 11 |
| 29 | Oct 17 | Seongnam v Suwon FC | 45% | 31% | 42% | 11 |
Two lessons in that table. First, every Suwon FC match matters to Seoul by roughly the same amount, 11 to 13 points of probability, because every one of them is a chance for Suwon FC to drop two or three points. There is no single decisive Suwon FC fixture; there are nine equally important ones. Second, Suwon Samsung’s other seven matches barely register for Seoul (none makes the top eight), because by the model’s reckoning Suwon Samsung are going up either way and the race for the second place runs between Seoul and Suwon FC.
The Cheongju match on the final day sits high on the list not because it is hard for Suwon FC (they win it 73 percent of the time) but because it comes last: if the race is still alive on November 29, that is the match that settles it, and a Cheongju side that has drawn 14 of 24 is exactly the kind of opponent that produces a 1-1.
Scenario five: Seoul’s own matches, ranked by what they are worth
The same exercise for Seoul’s fixtures. The final column is the difference in top-two probability between winning and not winning.
| Round | Opponent | Win chance | Top two if Seoul win | Top two if not | Swing |
|---|---|---|---|---|---|
| 26 | Suwon Samsung (H) | 38% | 55% | 26% | 29 pts |
| 27 | Daegu (H) | 47% | 49% | 26% | 23 |
| 31 | Yongin (A) | 43% | 49% | 28% | 21 |
| 34 | Gyeongnam (A) | 43% | 49% | 28% | 21 |
| 28 | Gimpo (A) | 54% | 46% | 26% | 20 |
| 33 | Chungnam Asan (H) | 59% | 45% | 26% | 19 |
| 32 | Jeonnam (H) | 68% | 43% | 25% | 18 |
| 29 | Gimhae (H) | 83% | 40% | 23% | 17 |
Saturday matters most, for the double reason that it is worth three points and takes points from a rival. After that the ranking is almost the reverse of difficulty: the matches Seoul are expected to win (Gimhae, Jeonnam) move the needle least when they are won, because the model already assumed it, and the coin flips (Daegu, Yongin, Gyeongnam) move it most. The model’s 4-3-1 sheet has Seoul taking one point from each of Daegu and Gyeongnam and none from Yongin. Turn any one of those into a win and the projection climbs two points and the top-two probability by roughly 20.
Scenario six: the extremes
- Seoul win out. Eight wins, 69 points, happens in about one simulation in 200. When it does, Seoul go up every time and win the title 94 percent of the time. It is not the plan; it is the ceiling.
- Seoul take 20 from 24 (six wins, two draws, 65 points). Top two 91 percent of the time. This is the realistic “we did everything right” season.
- Suwon Samsung collapse. They finish outside the top two 14 percent of the time. When they do, the pair going up is nearly always Seoul and Suwon FC (11 percent of all seasons). The route runs through Mokdong on Saturday and then Daegu and Seongnam away in late October.
- Daegu gatecrash. Daegu take one of the two places 7 percent of the time. It needs Edgar back, Cesinha fit, and two of the three above them to fall over. They visit Mokdong on September 19 and still have both Suwon clubs to play.
- Both Suwons up, Seoul to the playoffs. The single most likely outcome at 56 percent. Seoul would then be third in 43 percent of all simulations, which means a home semifinal on December 3 against whoever finishes sixth, most likely Busan or Hwaseong, and a final on December 6.
What a Seoul supporter should actually watch for
- Saturday, 16:30, Mokdong. Worth about 30 points of top-two probability on its own.
- Sunday, 19:00, Suwon Sports Complex. Suwon FC against Cheonan, and specifically whether Frizzo is in the squad. Cheonan have drawn ten of 24.
- October 9, 14:00. Suwon FC against Hwaseong, the one Suwon FC match the model rates under 50 percent. Seoul kick off at Gimpo two and a half hours later knowing the result.
- October 17. Seongnam against Suwon FC, the other sub-50 fixture. Seoul host Gimhae the next day.
- The scoreboard, not the table. Because the tiebreak is goals scored, a 3-0 for Seoul is worth more than a 1-0 in a way it would not be under a goal-difference rule, and every Suwon FC goal against Ansan and Cheongju widens a gap Seoul may need to close.
Caveats
The model does not know about injuries or suspensions that have not happened yet, and five Suwon FC regulars are one booking from missing a match. It treats Frizzo as absent for Round 26 only. It does not simulate positions below eighth, which does not affect the top-two question. Fixture dates for the Suwon clubs from Round 27 onward come from Sofascore and BeSoccer rather than the K League’s own schedule page, which does not render. And it is a model: Suwon FC were relegated last December after losing a playoff to the third-placed K League 2 side, the first time that had ever happened, which is a reminder that 56 percent is not 100.