Different result from the models

[AK140901]

------------------fanduel odds data------------------
Philadelphia 76ers (-124) @ Brooklyn Nets (106)
Phoenix Suns (-310) @ LA Clippers (250)
Milwaukee Bucks (-205) @ Miami Heat (172)
Memphis Grizzlies (160) @ Los Angeles Lakers (-190)
---------------XGBoost Model Predictions---------------
Brooklyn Nets vs Philadelphia 76ers (53.5%): UNDER 208.5 (73.6%)
LA Clippers (56.0%) vs Phoenix Suns: UNDER 226 (59.7%)
Miami Heat vs Milwaukee Bucks (52.6%): OVER 220.5 (50.4%)
Los Angeles Lakers vs Memphis Grizzlies (52.3%): UNDER 221 (61.8%)
--------------------Expected Value---------------------
Brooklyn Nets EV: -4.12
Philadelphia 76ers EV: -3.43
LA Clippers EV: 95.9
Phoenix Suns EV: -41.77
Miami Heat EV: 28.87
Milwaukee Bucks EV: -21.71
Los Angeles Lakers EV: -27.12
Memphis Grizzlies EV: 35.86

------------Neural Network Model Predictions-----------
1/1 [==============================] - 0s 196ms/step
1/1 [==============================] - 0s 20ms/step
1/1 [==============================] - 0s 16ms/step
1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 47ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 20ms/step
1/1 [==============================] - 0s 20ms/step
Brooklyn Nets (53.5%) vs Philadelphia 76ers: UNDER 208.5 (52.6%)
LA Clippers (58.2%) vs Phoenix Suns: OVER 226 (52.4%)
Miami Heat vs Milwaukee Bucks (58.0%): OVER 220.5 (53.4%)
Los Angeles Lakers (50.7%) vs Memphis Grizzlies: OVER 221 (54.2%)
--------------------Expected Value---------------------
Brooklyn Nets EV: 10.13
Philadelphia 76ers EV: -15.93
LA Clippers EV: 103.61
Phoenix Suns EV: -44.68
Miami Heat EV: 14.35
Milwaukee Bucks EV: -13.76
Los Angeles Lakers EV: -22.65
Memphis Grizzlies EV: 28.24

I ran both models today and in the first game between the Brooklyn Nets and the Philadelphia 76ers, the XGBoost model gave me a different result than the NN model. Can this be correct and which model is more reliable?

[kyleskom]

Why would there be different models and expecting them to give the same results?

[AK140901]

Can you explain me where are the differents between the models?

[chriseling]

One it’s based on xgboost and one looks like a regular neural network trained with back propagation

On Sat, Apr 22, 2023 at 2:17 PM AK140901 ***@***.***> wrote: Can you explain me where are the differents between the models? — Reply to this email directly, view it on GitHub <#207 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAVP5W72FA74L2BAQ6DK3BLXCRDGPANCNFSM6AAAAAAXHZ4HKM> . You are receiving this because you are subscribed to this thread.Message ID: <kyleskom/NBA-Machine-Learning-Sports-Betting/repo-discussions/207/comments/5697070 @github.com>

-- Best, Chris

[AK140901]

which one is better, or to put it another way, which model is it where 75% on the money line is promised?

[chriseling]

Nothing is gauranteed. The xgboost is 68.6% money line when it was validated but it could be different now.

On Sat, Apr 22, 2023 at 11:57 PM AK140901 ***@***.***> wrote: which one is better, or to put it another way, which model is it where 75% on the money line is promised? — Reply to this email directly, view it on GitHub <#207 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAVP5W7OM7COCWYEWBG3JELXCTHERANCNFSM6AAAAAAXHZ4HKM> . You are receiving this because you commented.Message ID: <kyleskom/NBA-Machine-Learning-Sports-Betting/repo-discussions/207/comments/5698465 @github.com>

-- Best, Chris