Confusion about map balance

I encountered a problem in my work today, and I was confused about the two solutions.

The problem is:
A matching game is assumed to be a world of LOL, Dota, or tanks.
On a map, there are 10000 sample battles. The team at the birth of a won 4700, the team at the birth of B won 4500, And the draw was 800;
Q: Is there a balance problem in this map?
If one party is biased towards the other, it may be a match problem or a map design problem;

 

The solution proposed by BI:
* Pick out 9200 non-draw battles
* Mark each game. If 0 wins, 1 wins, and 1 wins;
* Obtain a sequence of 0 and 1.
* Verify that "the mean value of the sample is 0.5 (both sides are balanced)" and t-test is used;

 

My proposed solution:

* Calculate the winning rate of each player on this chart;
* Example: Player 1 wins at point A: B wins: Draw is 37%: 35%: 28%; Player 2 wins at point A: B wins: Draw is 48%: 50%: 2%;
* The Dual Sample sequence is obtained based on the winning/losing ratio of each player;
* Perform a two-sample t-test;

Are the two conclusions Equal?

 

Comment:

The sequence 0, 1 in solution 1 does not conform to the normal distribution. You cannot directly use the t-test. In solution 2, the player winning rate is a normal distribution, which can be considered;