Mixed up engine

Why does, for move 15, the little arrow recommend Ne5, (which I in fact did) and meanwhile the stockfish analysis calls it an inaccuracy? Is the engine bi-polar?

I don't know I've had positions where stockfish recommends some others moves with the obvious checkmate move

@saurabhsangeeth
That's kind of hard to believe. But clearly there are two different analyses going on. The arrow recommendation and the evaluation appear to be different programs that don't always agree. That raises the further question of which is better.

@Zubbubu said in #4:
> @saurabhsangeeth
> That's kind of hard to believe. But clearly there are two different analyses going on. The arrow recommendation and the evaluation appear to be different programs that don't always agree. That raises the further question of which is better.
See it for yourself

The server analysis only goes 3,500,000 nodes deep (correct me if this number is wrong).
Otherwise server analyses would take forever.

It's an often reported harmless bug. In fact it actually provides useful information.

According to my local browser analysis, Ne3 is only the third best move. Qb3 is 0.4 centipawns better, so Ne5 is an inaccuracy for both engines.

@saurabhsangeeth

But I see the arrow recommending Rxd1#, so I'm not getting the problem.

@Zubbubu said in #9:
> @saurabhsangeeth
>
> But I see the arrow recommending Rxd1#, so I'm not getting the problem.
There are 2 other arrows

@StingerPuzzles said in #7:
> It's an often reported harmless bug. In fact it actually provides useful information.

Pray tell: what information is that?!