-- init schema & data
drop table if exists UserResult
create table UserResult(
	dt datetime,
	result int,
	[user] varchar(20)
)

insert into UserResult
select * from (values ('20220917 12:00:00', 1,'a')
,('20220917 10:10:00',1,'a')
,('20220917 11:05:00',1,'a')
,('20220917 15:08:00',0,'a')
,('20220917 16:01:00',0,'a')
,('20220917 17:00:00',1,'a')
,('20220917 18:00:00',1,'a') 
) as temp(dt, result, [user])

-- result here
;with temp as(
	select result, 
		dt,
		[user],
		Row_Number() over(order by dt) as rid,
		ROW_NUMBER() OVER (PARTITION BY result ORDER BY dt) AS rn    
	from UserResult
)
select top 1 result, count(*) as number
from temp
group by result, rid - rn
order by count(*) desc

參考資料

如何對連續紀錄做分群