Skip to content
mape tournament
Español

Run e4af42be5992 · 30 origin dates · 6 room nights · 9 models · the hotel

Who Forecast Each Day Best?

Every model forecasts tomorrow’s room nights, breakfasts and parking spaces from the days before it. Each tile is that day’s MAPE: lower is better, and the tomato corner marks the winner.

SARIMAX + weather won the run with an average MAPE of 34.5, 6.0 points below the naive baseline, Same weekday.

Have sales of your own? Run the same tournament on them — in your browser, and the file never leaves it.

MAPE per model per origin date
Dec 2025–Jan 2026Daysame-weekdayensemble-v2thetaholt-winterssarimaxrandom-forestxgboostprophetlstm
Sun 7
37.2
40.0
51.8
38.1
34.6
38.7
38.2
33.7(winner)
52.6
Mon 8
33.6(winner)
39.1
48.2
37.9
40.1
42.4
47.1
50.2
66.0
Tue 9
34.5
38.4
43.4
24.2
24.1(winner)
30.4
27.7
25.4
53.5
Wed 10
50.4
44.3
38.1(winner)
46.1
47.3
46.2
50.7
45.8
49.3
Thu 11
16.7
15.3
18.4
15.3
15.0(winner)
15.8
15.7
19.9
22.2
Fri 12
15.7
17.2
20.5
14.8
15.7
14.8
14.7
12.7(winner)
23.6
Sat 13
11.8
13.8
21.4
11.1
10.5(winner)
12.2
16.5
12.3
22.9
Sun 14
37.4
49.0
62.3
36.2
34.8
39.6
38.1
34.2(winner)
73.8
Mon 15
40.5(winner)
55.7
73.0
48.9
44.8
53.4
48.4
51.7
84.0
Tue 16
22.4
26.0
34.9
20.0
15.5(winner)
29.8
23.8
18.7
43.5
Wed 17
27.0
25.3
26.2
13.7
13.6(winner)
24.6
23.2
13.9
37.3
Thu 18
26.6
27.7
28.7
25.2
25.1(winner)
26.9
27.8
25.6
27.7
Fri 19
22.2
23.1
23.1
16.9
16.1
14.7
14.3(winner)
18.5
19.7
Sat 20
25.6
32.8
43.7
16.2(winner)
27.4
23.9
21.0
17.3
50.8
Sun 21
39.6
46.6
59.6
34.0
27.5
24.9
20.6(winner)
39.6
71.0
Mon 22
74.8
83.0
103.6
80.3
65.8(winner)
82.2
70.1
80.8
130.1
Tue 23
135.5
132.5
127.8
110.1
88.7
74.0(winner)
96.0
116.0
153.1
Wed 24
41.3
39.4
40.5
32.2
14.7(winner)
25.8
18.9
31.8
58.9
Thu 25
65.3
61.6
61.3
66.6
54.8(winner)
58.2
67.7
65.5
81.4
Fri 26
72.9
64.4
53.9(winner)
70.7
57.3
70.0
71.9
65.3
65.5
Sat 27
28.5(winner)
36.2
46.7
37.6
30.0
36.9
36.5
35.8
52.7
Sun 28
46.0
62.2
81.9
43.6
38.9
46.2
44.7
38.7(winner)
96.6
Mon 29
24.9
22.6
36.0
25.4
25.6
18.4(winner)
19.9
25.8
43.5
Tue 30
25.3
16.9(winner)
17.6
20.9
18.2
26.4
25.2
23.8
26.8
Wed 31
15.5
9.4
10.4
7.7
6.8(winner)
11.4
15.1
14.5
31.2
Thu 1
29.3
31.0
40.2
39.5
35.9
27.5(winner)
30.9
38.9
43.9
Fri 2
102.6
82.0
71.3(winner)
86.8
81.7
91.6
88.1
94.0
88.4
Sat 3
23.4
16.7
20.3
20.2
21.3
18.0
17.0
16.6
12.1(winner)
Sun 4
71.8
70.7
90.8
75.5
72.9
82.3
58.3(winner)
90.6
119.7
Mon 5
15.2
15.0
14.7
12.9
30.2
27.2
30.9
9.7(winner)
28.9

SW same-weekday · ENS ensemble-v2 · TH theta · HW holt-winters · SAR sarimax · RF random-forest · XGB xgboost · prophet · lstm

Scroll the table for all 30 days — the model row and the day column stay in view.

MAPE %lower is better

  • 0–20
  • 20–25
  • 25–30
  • 30–40
  • 40+
  • ND
  • fail
  • short
  • day winner

Best average

34.5 %

sarimax

Steadiest

σ 21.1

sarimax

Leaderboard · avg MAPElower is better

  1. 01SARIMAX + weathersarimax34.5σ 21.1 · 30 d
  2. 02XGBoostxgboost37.3σ 22.3 · 30 d
  3. 03Holt-Wintersholt-winters37.6σ 25.3 · 30 d
  4. 04Random Forestrandom-forest37.8σ 22.3 · 30 d
  5. 05Prophetprophet38.9σ 26.9 · 30 d
  6. 06Same weekdaysame-weekday40.5σ 27.4 · 30 d
  7. 07Ensemble v2ensemble-v241.3σ 26.3 · 30 d
  8. 08Thetatheta47.0σ 27.5 · 30 d
  9. 09LSTMlstm57.7σ 33.8 · 30 d

Order · run e4af42be599211 Oct 2026

Dataset
hotel-year · seed 12 · 6da9…
Window
30 origin days · h = 7 · forecast 7 days ahead
Roster
9 models · 3 light, 6 heavy
Engine
py 3.12.15 · uv.lock b7df…
Result hash
80b1 011e 49f5