My guess is that Chrome increases the site engagement score as a result of you interacting with the website.
1. When the object is returned, go to chrome://site-engagement - is it still below 30?
2. When the object is returned, is Flash still blocked?

PFB replies

1) We know chrome increases site engagement. But we went and explicitly changed it to less than 30. When object was returned, we checked site-engagement it was less than 30. We checked with values like 8 and 25, 28, etc. 

2) When object is returned, chrome displayed a grey clickable screen. So flash is still blocked by chrome and our game didn't load.

Above issue happened during 12 to 1 PM IST today. 
After that we're unable to reproduce. Chrome is not having that object when flash is blocked.

Hi Phistuck,

We've got the exact reproduction steps for above issue. Above issue is happening once site engagement goes above 30 and we manually reset it to a value less than 30. At that point in time, if you check for navigator.plugins it contains 'shockwave flash' object, but flash is blocked by chrome

If you restart chrome, navigator.plugins is getting reset and doesn't contain shockwave flash object.

We tried using navigator.mimetypes also, but same behavior is happening for that property too.

Please help us with below two questions.
1) Is this issue with modifying site-engagement score? 
2) I know site engagement score increases with each visit. But let's say, I didn't visit a particular site for long time. Will my previous site engagement go down?

Exact steps to reproduce:

1) Take a flash website, goto site-engagement and modify to a score less than 30 (say 25).
2) Visit the site multiple times and check for navigator.plugins or navigator.mimeTypes to contain 'shockwave flash' object. It won't contain the object.
3) Visit the site multiple times, so that site-engagement score becomes greater than 30
4) At this point, goto the same flash site, check for navigator.plugin or navigator.mimeTypes. Both will contain 'shockwave flash' object. But Flash would be blocked by chrome.
5) Now restart chrome without any changes. Goto same flash site, navigator.plugins or navigator.mimeTypes would not contain 'shockwave flash'

Expected behavior:
Whenever flash is blocked by chrome, navigator.plugin or navigator.mimeTypes should not contain 'shockwave flash' object.

Step 5 seems a bit unexpected as well, since the site engagement score at this point is larger than 30 (see step 3), but Flash still does not exist.

Oops. My bad! missed a step. Below is correct steps.

Exact steps to reproduce:

1) Take a flash website, goto site-engagement and modify to a score less than 30 (say 25).
2) Visit the site multiple times and check for navigator.plugins or navigator.mimeTypes to contain 'shockwave flash' object. It won't contain the object.
3) Visit the site multiple times, so that site-engagement score becomes greater than 30
4) Now goto site-engagement and manually change the score to a value lesser than 30.
5) At this point, goto the same flash site, check for navigator.plugin or navigator.mimeTypes. Both will contain 'shockwave flash' object. But Flash would be blocked by chrome.
6) Now restart chrome without any changes. Goto same flash site, navigator.plugins or navigator.mimeTypes would not contain 'shockwave flash'

I am trying to reproduce the bug, but added stable blocker for now.

Removing the release block label.  Agreed that this is an odd state that we should look into, however given that manual manipulation of the site engagement score is required to reproduce this issue, it likely isn't a common case that most users would discover (be able to discover to do).

Tested the same on mac 10.11.6 chrome version 56.0.2924.3 - navigator.plugins['Shockwave Flash']  in console displays undefined

Confirming the issue for further inputs on this

Hey - Dom, Raymes, and I are aware of the inconsistent behavior that can occur when site engagement changes.

We've discussed the possibility of making the SEI score 'sticky' per navigation, but none of us have considered it a high enough priority to actually implement this. And until this bug report, no one had actually hit this edge case in the wild.

In all cases, I believe navigator.plugins reflects the actual state of the plugin list. If the SEI score increases, and Flash then appears in navigator.plugins, subsequent attempts to instantiate a Flash object will succeed (not be blocked).