At the end of Episode 13, Chen Hui’s gaze turns steely, revealing the moment he realizes his situation has become dangerous, possibly crossing the point of no return.
The life he ventured into in the city was full of bad characters. Young, naïve, inexperienced, and unaware of the perils around him, he never stood a chance. Given his desperate situation, it was inevitable that he would fall prey to them.
Songge always believed in Chen Hui because he was the only person who ever truly protected her. Some people say she’s selfish for letting him do unlawful things, but from her perspective, when he says he won’t let her down, she trusts him completely. That’s why she follows him no matter what he does — her loyalty comes from feeling safe with him in a way she never felt with anyone else.
Chen Hui was not the first one in his family to resort to illegal activity to help out the family expenses. After they learned of her illness, his mom treated people in an illegal coal pit to earn more money. The dad seems like he never really tried hard to earn more money because Chen Hui's mom said he had a measly salary that forced her to sell her seniority for payout (actually she got fired probably for doing extra work at illegal coal mine treating patients there) and work in illegal coal mine, and Chen Hui and Songge could hear them arguing over finances, think how Songge must feel. Chen Hui's dad doesn't try to do anything, he always just says things like, "I'll just sell the house," or "I'll just quit my job." He's really only concern about his own thing. It's not that he doesn't love his family, but he's very selfish. His wife had to take financial matters into her own hands. Her husband doesn't seem to act. It's her payout money that is sustaining them in the meantime, but it will run out too. That's why Chen Hui had to leave home to find a way to earn money.
The husband is all self-righteous about what is right and what is wrong, but the wife is willing to cross a line while the husband is not doing anything about their dire financial situation. So, his inaction is forcing others in his family to take desperate measures into their own hands.
The husband attitude is that he will handle things, but he doesn't talk to his family and yet he isn't doing anything about it, so how do they know what he's going to do? They don't. He's the kind person who doesn't' communicate with his family. Some people are like that, they think you don't need to know, I'll take care of it, but this is not a matter you can just everyday don't do anything. I think he was going to sell the house, but then where would they live? He really has no solution and he's not addressing it nor communicating it. The communication breakdown in the family is really him.
This type of dad pushes his son away from him. Most parents like that don't realize it until the relationship sours, some never recognize why. Parents don't realize when your kid is around 18, they can essentially leave home and try to support themselves. on their own, and if you never had a good relationship with them, they may never come back to see you. But Chen Hui wont' do that because he wants to save Songge.
At the very start of the intro, a circle—800 meters in radius—was drawn on a map. Then a shockwave rippled across it like an earthquake. Some kind of black medicinal powder spilled out, landed on a fuse, ignited, and set everything on fire.
How did all these pushes and pulls — the 18 highs, the 13 lows, and the CW rating drop — cancel out so perfectly that the global score stayed at 7.3? At this point the pattern really does look like MDL is smoothing or correcting the data, which is why the posted scores don’t feel trustworthy.
Start: 355 user ratings at 7.3. 18 new ratings at 10 → +0.13. 13 new ratings at 1 → −0.22. Net from new ratings = −0.09 (total now 386 = 355+18+13). The 18 highs come from Completed increasing by 16, explained by 17 CW users moving to Completed plus 1 brand‑new Completed rating. The 13 lows come from the drop‑user count increasing by 13, and their average falling from 2.31→1.97 — meaning 13 new drop‑users rated around 1. Despite adding 18 very high scores (>7.23), the Completed average barely moves (7.23→7.22). The upward impact of 18 highs is so large that only broad‑based downgrades or a cluster of major downward edits could overpower it, making the slight drop behaviorally odd. How did all these pushes and pulls — the 18 highs, the 13 lows, and the CW rating drop — cancel out so perfectly that the global score stayed at 7.3? If nothing else changed, the global score should fall to ~7.21 with 386 ratings. To keep the global score at 7.3, the 90 existing CW raters (total user ratings = 386, and 296 of those are in Completed, leaving 90 ratings coming from CW) would need to raise their ratings by about +0.4 each, adding back +0.09. But the overall CW average (all 529 CW users) fell from 7.88→7.63. For that to be true while the 90 went up, the remaining 439 CW users would need to drop their ratings by about −0.38 each. So the only way to make all numbers fit is: 90 CW raters go UP ~0.4, 439 CW users go DOWN ~0.38, CW average still drops 0.25, Completed barely changes, and the global score stays at 7.3 — mathematically possible, but behaviorally very unnatural.
The main reason the score sits at 7.3 is that 20–25% of all ratings are 1-stars. This is a very heavy anchor. Even if most people are giving it 9s and 10s, that many 1-star ratings drag the overall average down hard and keeps it stuck around 7.3. If people are using multiple accounts, then the 1‑stars may be coming from a small group — like 10 to 15 users.
But MDL is still manipulating the final result by smoothing the numbers, keeping the main score locked at 7.3. The score is influenced by real ratings, but MDL is actively intervening to make it look stable and clean.
are you a statistician 😆😆 you seem to like numbers 😆 when i read your comments about the ratings I can…
I can't stand data that makes no sense. You can't drop the Currently Watching to 7.63 unless you remove 72 perfect 10 scores from it and then it has to move to the completed rating and also impact the overall score by +0.4. That's what the math works out to.
It feels like the score is dropping to 7.1 or going back to 7.5. What's with the big drops of 1 stars every time…
Dropped got another boat load 18 new 1‑stars, which is why it fell from 2.31 → 2.09 — that alone has a −0.3 impact on the drama’s total score. Completed got 11 new ratings, 10 of them coming from Currently Watching — that should raise the Completed score, not lower it, especially since CW went down. But even if all 10 CW leavers were perfect 10s, CW should only drop from 7.94 → ~7.89 but CW is showing 7.63, which is mathematically impossible — you’d need another 62 perfect‑10 users (72 total) to leave CW to reach that number, and those 62 would also have to appear in Completed (they don’t). Either the data isn’t aligned yet, or the scoring logic is broken.
Let's assume the CW wasn’t updated yet, just the 18 new 1‑stars and the 11 perfect‑10s would move the total score from 7.3 → 7.08 or 7.1.
It feels like the score is dropping to 7.1 or going back to 7.5. What's with the big drops of 1 stars every time there may be 10 stars
Dropped got another boat load 18 new 1‑stars, which is why it fell from 2.31 → 2.09 — that alone has a −0.3 impact on the drama’s total score. But the Currently Watching can only drop to 7.63 if there are 72 user ratings at 10-stars. If this is real, it will have +0.5 impact on the score
Douban currently gives Eight Hundred a 6.9, which feels heavily influenced by the past 2 days’ ratings. Based…
I took over a dozen random samples from 4/27 to 5/6, each with around 100 comments. The average kept bouncing between 6.9 and 7.1. In the latest comments, 5-star ratings became less frequent while 3s and 4s increased, and 1-stars clearly decreased (though I didn’t sample these most recent ones). I’ve never seen a Xu Kai drama flooded with fake 5-stars, which makes Douban comments more usable compared to some other popular dramas that are flooded with perfect 5-stars. Because the comments feel genuine, you can actually sample and analyze the ratings properly. From my sampling, I mostly saw 7.0–7.1, with some 6.9s and one 6.5. The score had been trending upward (it even hit 7.6 at one point), so I stopped checking when it stabilized around 7.1. But it seems the past two days’ comments shifted again, lower but I did not take samples. I think it would have shown some 6.9 again. It really feels like the current rating is heavily influenced by the most recent votes. Overall, I still think Douban is far more credible than MDL. At least Douban waits until the score looks somewhat stable before displaying it. That said, it has issues too — some dramas buy 5-stars for themselves and 1-stars to attack competing dramas.
Douban currently gives Eight Hundred a 6.9, which feels heavily influenced by the past 2 days’ ratings. Based on my sampling, they could have easily rounded it to 7. The last couple of days showed more 3- and 4-star ratings than 5-stars. At least the early 1-star reviews dropped sharply from about 10% to 3% (unlike on MDL where 1-stars seem to be 20-25%).
If votes are mixed, Douban becomes overly cautious, so the score lands on the low end. But if it’s mostly perfect 5-star votes, they err on the high end — they can’t remove them all.
It looks like—whoever they are—impeccable timing came a day before Douban score and started dropping 3‑star ratings. It's like they know when to do it to make it look real. "Why is it so low, after reading the short reviews, there were suddenly a lot of low-scoring and bad reviews poured in today."
In my dozen samples, the 5 stars mix was mostly above 20% except couple of times it dipped to 18%, but my last sample on 5/6, it was 17% So it sure feels like 6.9 is heavily influenced by the recent ratings. In fact, my last sample on 5/6 almost mirrors Douban mix except that I saw way more 4 stars than 3 stars in that instant. I thought it was pretty stable so I stopped sampling but apparently there must be a lot of 3 stars and Douban decided to go with recent ratings. It's no wonder the dramas that are flooded with perfect 5 stars don't let up not even long after the drama has aired.
Today, 10 new user ratings were added. About 2 users who had previously dropped the show but never rated it came back to finish watching the finale and gave it scores in the 5–7 range (which is reasonable for a drop user), but they remained in the drop status. This explains how the drop rating increased from 2.16 to 2.31.
Around 8 users moved from “Currently Watching” to “Completed” and rated it for the first time, avg score 9.5. This explains how the completed rating increased from 7.16 to 7.23, and why it’s difficult to push it back toward an 8. The impact wasn’t strong enough to move the overall drama score, which remained at 7.3.
When the completed rating finally rose to 7.93 on 5/6, 18 one‑star ratings from the drop status were moved into the completed category, bringing the completed score down to its current level.
Xu Kai is in the “final contract stretch” period. I think he's just waiting for it to naturally expire on its own. This is the cleanest and lowest risk path. But that becomes a stalemate with his agency. Once his contract expires, I think his agency will create negativity about Xu Kai so no one will sign with him. Ideally, you want a popular drama airing as your contract expires, but his agency won't support him if he's not renewing.
I have two data points on 5/6, 8am and 5pm. My conclusion is 18 1-star raters moved their rating from the drop status to the completed status and only impacted the completed rating, not the overall score because they are already baked in 15 new user ratings in the completed avg rating 9.32 33 accounts for the total positive change in the completed status it validates how the drop rating improved from 1.96 to 2.16 and how the completed rating declined from 7.68 to 7.32
There are 27 new user ratings (cannot include the 18 1-star raters who were not new) 15 came from the completed avg rating 9.32 above 12 came from plan to watch avg rating 2.2 <- this is strange but it's the only bucket left this scenario would drop the drama score from 7.4 to 7.3 which it did
now at 8pm the completed rating declined further to 7.16 9 out of the 12 above with rating 2.2 moved from plan to watch to the completed status and if 3 new user completions at rating of 9 this accounts for the total positive change of 12 in completed status this scenario would drop it to 7.16 the drama score stays at 7.3 because the 9 at 2.2 was already baked in
MDL the score pattern can never be higher than the completed rating so if it doesn't change or keeps declining the drama score will go below 7.1
The data strongly suggests that the finale brought a binge wave on 5/5/26 (people who jumped straight into Completed who were never in the Currently Watching status or any other status) rated the show well below 8.
When the Completed average dropped from ~7.93 → 7.68 with +26 new completers, the implied average rating of those specific 26 people is approximately 5.8 – 6.2.
now the drama has ended, it's still spreading by word of mouth organically, and people who were not interested are having second thoughts because of the great reviews out there
Are there no other actors taking time off except Xu Kai?
Xu Kai made trade‑offs to take on less popular projects with much lower pay and fewer endorsements, like The Eight Hundred and Fire Investigator. That shows that money and popularity are less important to him right now than exploring new roles, and it shows he’s willing to take some risks. These choices will open the door to more non‑idol projects.
2 
1 
The husband is all self-righteous about what is right and what is wrong, but the wife is willing to cross a line while the husband is not doing anything about their dire financial situation. So, his inaction is forcing others in his family to take desperate measures into their own hands.
The husband attitude is that he will handle things, but he doesn't talk to his family and yet he isn't doing anything about it, so how do they know what he's going to do? They don't. He's the kind person who doesn't' communicate with his family. Some people are like that, they think you don't need to know, I'll take care of it, but this is not a matter you can just everyday don't do anything. I think he was going to sell the house, but then where would they live? He really has no solution and he's not addressing it nor communicating it. The communication breakdown in the family is really him.
This type of dad pushes his son away from him. Most parents like that don't realize it until the relationship sours, some never recognize why. Parents don't realize when your kid is around 18, they can essentially leave home and try to support themselves. on their own, and if you never had a good relationship with them, they may never come back to see you. But Chen Hui wont' do that because he wants to save Songge.
Start: 355 user ratings at 7.3.
18 new ratings at 10 → +0.13.
13 new ratings at 1 → −0.22.
Net from new ratings = −0.09 (total now 386 = 355+18+13).
The 18 highs come from Completed increasing by 16, explained by 17 CW users moving to Completed plus 1 brand‑new Completed rating.
The 13 lows come from the drop‑user count increasing by 13, and their average falling from 2.31→1.97 — meaning 13 new drop‑users rated around 1.
Despite adding 18 very high scores (>7.23), the Completed average barely moves (7.23→7.22). The upward impact of 18 highs is so large that only broad‑based downgrades or a cluster of major downward edits could overpower it, making the slight drop behaviorally odd.
How did all these pushes and pulls — the 18 highs, the 13 lows, and the CW rating drop — cancel out so perfectly that the global score stayed at 7.3?
If nothing else changed, the global score should fall to ~7.21 with 386 ratings.
To keep the global score at 7.3, the 90 existing CW raters (total user ratings = 386, and 296 of those are in Completed, leaving 90 ratings coming from CW) would need to raise their ratings by about +0.4 each, adding back +0.09.
But the overall CW average (all 529 CW users) fell from 7.88→7.63.
For that to be true while the 90 went up, the remaining 439 CW users would need to drop their ratings by about −0.38 each.
So the only way to make all numbers fit is: 90 CW raters go UP ~0.4, 439 CW users go DOWN ~0.38, CW average still drops 0.25, Completed barely changes, and the global score stays at 7.3 — mathematically possible, but behaviorally very unnatural.
The main reason the score sits at 7.3 is that 20–25% of all ratings are 1-stars.
This is a very heavy anchor. Even if most people are giving it 9s and 10s, that many 1-star ratings drag the overall average down hard and keeps it stuck around 7.3.
If people are using multiple accounts, then the 1‑stars may be coming from a small group — like 10 to 15 users.
But MDL is still manipulating the final result by smoothing the numbers, keeping the main score locked at 7.3. The score is influenced by real ratings, but MDL is actively intervening to make it look stable and clean.
Completed got 11 new ratings, 10 of them coming from Currently Watching — that should raise the Completed score, not lower it, especially since CW went down.
But even if all 10 CW leavers were perfect 10s, CW should only drop from 7.94 → ~7.89 but
CW is showing 7.63, which is mathematically impossible — you’d need another 62 perfect‑10 users (72 total) to leave CW to reach that number, and those 62 would also have to appear in Completed (they don’t).
Either the data isn’t aligned yet, or the scoring logic is broken.
Let's assume the CW wasn’t updated yet, just the 18 new 1‑stars and the 11 perfect‑10s would move the total score from 7.3 → 7.08 or 7.1.
Dropped got another boat load 18 new 1‑stars, which is why it fell from 2.31 → 2.09 — that alone has a −0.3 impact on the drama’s total score.
But the Currently Watching can only drop to 7.63 if there are 72 user ratings at 10-stars. If this is real, it will have +0.5 impact on the score
I’ve never seen a Xu Kai drama flooded with fake 5-stars, which makes Douban comments more usable compared to some other popular dramas that are flooded with perfect 5-stars. Because the comments feel genuine, you can actually sample and analyze the ratings properly.
From my sampling, I mostly saw 7.0–7.1, with some 6.9s and one 6.5. The score had been trending upward (it even hit 7.6 at one point), so I stopped checking when it stabilized around 7.1. But it seems the past two days’ comments shifted again, lower but I did not take samples. I think it would have shown some 6.9 again. It really feels like the current rating is heavily influenced by the most recent votes.
Overall, I still think Douban is far more credible than MDL. At least Douban waits until the score looks somewhat stable before displaying it. That said, it has issues too — some dramas buy 5-stars for themselves and 1-stars to attack competing dramas.
Score history I tracked:
4/27-4/28 7.2
4/29: 7.3
5/1: 6.9
5/3: 6.9
5/4: 7.6
5/5–5/6: 7- 7.1
5/7 likely 6.9 - 7
If votes are mixed, Douban becomes overly cautious, so the score lands on the low end. But if it’s mostly perfect 5-star votes, they err on the high end — they can’t remove them all.
It looks like—whoever they are—impeccable timing came a day before Douban score and started dropping 3‑star ratings. It's like they know when to do it to make it look real.
"Why is it so low, after reading the short reviews, there were suddenly a lot of low-scoring and bad reviews poured in today."
Douban 6.9
16,073 reviews
5★ — 17.1%
4★ — 31.5%
3★ — 35.4%
2★ — 12.8%
1★ — 3.2%
In my dozen samples, the 5 stars mix was mostly above 20% except couple of times it dipped to 18%, but my last sample on 5/6, it was 17% So it sure feels like 6.9 is heavily influenced by the recent ratings. In fact, my last sample on 5/6 almost mirrors Douban mix except that I saw way more 4 stars than 3 stars in that instant. I thought it was pretty stable so I stopped sampling but apparently there must be a lot of 3 stars and Douban decided to go with recent ratings. It's no wonder the dramas that are flooded with perfect 5 stars don't let up not even long after the drama has aired.
My 5/5 - 5/6 samples both 7.1
5 → 22% 5 → 17%
4 → 32% 4 → 39%
3 → 30% 3 → 29%
2 → 13% 2 → 12%
1 → 3% 1 → 3%
Around 8 users moved from “Currently Watching” to “Completed” and rated it for the first time, avg score 9.5. This explains how the completed rating increased from 7.16 to 7.23, and why it’s difficult to push it back toward an 8. The impact wasn’t strong enough to move the overall drama score, which remained at 7.3.
When the completed rating finally rose to 7.93 on 5/6, 18 one‑star ratings from the drop status were moved into the completed category, bringing the completed score down to its current level.
https://www.youtube.com/shorts/cXBEKn1bQAo?feature=share
18 1-star raters moved their rating from the drop status to the completed status and only impacted the completed rating, not the overall score because they are already baked in
15 new user ratings in the completed avg rating 9.32
33 accounts for the total positive change in the completed status
it validates how the drop rating improved from 1.96 to 2.16
and how the completed rating declined from 7.68 to 7.32
There are 27 new user ratings (cannot include the 18 1-star raters who were not new)
15 came from the completed avg rating 9.32 above
12 came from plan to watch avg rating 2.2 <- this is strange but it's the only bucket left
this scenario would drop the drama score from 7.4 to 7.3 which it did
now at 8pm the completed rating declined further to 7.16
9 out of the 12 above with rating 2.2 moved from plan to watch to the completed status
and if 3 new user completions at rating of 9
this accounts for the total positive change of 12 in completed status
this scenario would drop it to 7.16
the drama score stays at 7.3 because the 9 at 2.2 was already baked in
MDL the score pattern can never be higher than the completed rating so if it doesn't change or keeps declining the drama score will go below 7.1
When the Completed average dropped from ~7.93 → 7.68 with +26 new completers, the implied average rating of those specific 26 people is approximately 5.8 – 6.2.
Xu Kai made trade‑offs to take on less popular projects with much lower pay and fewer endorsements, like The Eight Hundred and Fire Investigator. That shows that money and popularity are less important to him right now than exploring new roles, and it shows he’s willing to take some risks. These choices will open the door to more non‑idol projects.