Page **2** of **2**

### Re: Scores

Posted: **January 30th, 2020, 11:20 am**

by **mjcox2000**

At MIT, the assertion that scores need to be very nearly perfect to place is mostly false. Multiple teams medaled with temperature scores near or below 52. Conversely, there were teams with temperature scores above 55 that placed in the teens and temperature scores above 50 that placed in the 20s.

If there's interest, I could compute correlation coefficients and such.

### Re: Scores

Posted: **January 30th, 2020, 1:04 pm**

by **YeagerTheCat**

So that only tells us that either they got nailed on the design log somehow (hard to see how the would screw that up badly) or more specifically they got hurt in the test. 28 points on the test, a device that gets a full 60 only gets 5 point advantage over 55 scoring device (I work in the department of the obvious

). A single question can knock that out asking about energy gaps in semiconductors. The short answer is you have to do well in everything and a great device does not buy you very much (arguably not enough).

### Re: Scores

Posted: **January 30th, 2020, 1:20 pm**

by **knightmoves**

YeagerTheCat wrote: ↑January 30th, 2020, 1:04 pm

28 points on the test, a device that gets a full 60 only gets 5 point advantage over 55 scoring device (I work in the department of the obvious

). A single question can knock that out asking about energy gaps in semiconductors. The short answer is you have to do well in everything and a great device does not buy you very much (arguably not enough).

The written test is worth 30 points. A 5 point gain on the device is 17% of the total points available from the test. You're right - you can't bomb the test and still win on the back of a good device, but having a 17% lead over another team going in to the test is a pretty substantial head start.

### Re: Scores

Posted: **January 30th, 2020, 1:46 pm**

by **YeagerTheCat**

Oh I would take it, no worries, just saying that it will be won or lost on the test for a lot of teams.

### Re: Scores

Posted: **January 30th, 2020, 2:20 pm**

by **LIPX3**

mjcox2000 wrote: ↑January 30th, 2020, 11:20 am

At MIT, the assertion that scores need to be very nearly perfect to place is mostly false. Multiple teams medaled with temperature scores near or below 52. Conversely, there were teams with temperature scores above 55 that placed in the teens and temperature scores above 50 that placed in the 20s.

If there's interest, I could compute correlation coefficients and such.

Did MIT use Nationals rules? If so, then 52 would be an average of 0.5 C off.

### Re: Scores

Posted: **January 30th, 2020, 5:00 pm**

by **Umaroth**

LIPX3 wrote: ↑January 30th, 2020, 2:20 pm

mjcox2000 wrote: ↑January 30th, 2020, 11:20 am

At MIT, the assertion that scores need to be very nearly perfect to place is mostly false. Multiple teams medaled with temperature scores near or below 52. Conversely, there were teams with temperature scores above 55 that placed in the teens and temperature scores above 50 that placed in the 20s.

If there's interest, I could compute correlation coefficients and such.

Did MIT use Nationals rules? If so, then 52 would be an average of 0.5 C off.

MIT did use nationals rules

### Re: Scores

Posted: **January 30th, 2020, 7:04 pm**

by **YeagerTheCat**

National rules are absolutely a lot harder than matching your own thermometer.

### Re: Scores

Posted: **January 31st, 2020, 6:01 am**

by **LIPX3**

YeagerTheCat wrote: ↑January 30th, 2020, 7:04 pm

National rules are absolutely a lot harder than matching your own thermometer.

Then people saying you don't need to be perfect are misunderstanding what is being said - those peformances at MIT required to medal are equivalent to perfect or near perfect performances at Regionals and State.