สำเนาบทสนทนา 🔗
WEBVTT
00:00:02.080 --> 00:00:03.200
0.2.
00:00:03.520 --> 00:00:14.160
That is the gap in percentage points between the highest and lowest of Metro North's three big lines over the last 12 months, according to the MTA's own data.
00:00:14.400 --> 00:00:19.359
Every morning, somebody on a Westchester platform swears their line is the best one.
00:00:19.519 --> 00:00:21.920
Every morning, the spreadsheet is less certain.
00:00:22.079 --> 00:00:25.280
Hudson, Harlem, or New Haven, which line is the best?
00:00:25.519 --> 00:00:27.199
This is the Friday Intel.
00:00:27.519 --> 00:00:29.039
Let's go deeper.
00:00:30.079 --> 00:00:33.439
Welcome to the Friday Intel from I Live Here Westchester.
00:00:33.600 --> 00:00:39.359
Every Friday, we go deeper on one data story that affects your life in Westchester County this week.
00:00:39.600 --> 00:00:42.640
Which Metro North Line is the best for commuters?
00:00:42.799 --> 00:00:48.799
Last month we weighed New York against New Haven on pizza, and both sides turned out to be right about different things.
00:00:49.039 --> 00:00:52.240
New Haven is back, this time as a railroad.
00:00:52.399 --> 00:00:53.840
The city has range.
00:00:54.000 --> 00:01:00.399
Our source is the MTA's own on-time performance data set on the state open data portal.
00:01:00.560 --> 00:01:06.400
It counts the share of trains that reach their final stop within five minutes and 59 seconds of schedule.
00:01:06.560 --> 00:01:11.920
We average the monthly scores from October 2025 through September 2026.
00:01:12.640 --> 00:01:17.760
Here are the averages, the Hudson Line, 97.4% on time.
00:01:18.079 --> 00:01:23.040
New Haven, 97.3%, Harlem, 97.2%.
00:01:23.599 --> 00:01:25.200
That is the whole race.
00:01:25.439 --> 00:01:29.120
Harlem and New Haven are four hundredths of a point apart.
00:01:29.359 --> 00:01:33.040
The Hudson Line has the river, the New Haven line has Connecticut.
00:01:33.200 --> 00:01:36.799
The Harlem Line has Wassack, which is fun to say.
00:01:37.280 --> 00:01:38.400
The fine print.
00:01:38.640 --> 00:01:41.519
These are whole line numbers, not Westchester numbers.
00:01:41.760 --> 00:01:44.000
Each line runs well past the county.
00:01:44.159 --> 00:01:49.359
The Hudson to Poughkeepsie, the Harlem to Wasaic, the New Haven into Connecticut.
00:01:49.599 --> 00:01:57.760
Nobody sliced out the Westchester stops, and it is a plain average of twelve monthly scores with no waiting for how many trains ran.
00:01:59.439 --> 00:02:04.159
Month by month, the lines trade places like commuters swapping seats.
00:02:04.319 --> 00:02:06.959
Hudson finished on top in four months.
00:02:07.200 --> 00:02:10.560
New Haven, four, Harlem, three.
00:02:10.960 --> 00:02:17.840
In September, Harlem and New Haven tied at ninety-seven point eight, with Hudson at ninety-seven point seven.
00:02:18.000 --> 00:02:20.879
A simple month-by-month comparison we ran ourselves.
00:02:21.039 --> 00:02:27.599
Our arithmetic and not the MTAs found gaps too small to separate from ordinary monthly noise.
00:02:27.759 --> 00:02:31.280
On overall on time rate, the lines cannot be separated.
00:02:31.520 --> 00:02:36.479
Split by rush hour, and the gaps widened, though still by a small margin.
00:02:36.639 --> 00:02:44.800
On morning peak trains, Hudson averaged 95.5, Harlem 95.1, and New Haven 94.6.
00:02:45.280 --> 00:02:47.039
In the evening, it flips.
00:02:47.280 --> 00:02:58.000
New Haven leads at ninety-eight, about nine-tenths of a point ahead of Hudson at ninety seven point one, and one and a half ahead of Harlem at ninety six point five.
00:02:58.240 --> 00:03:01.039
New Haven is last in the morning and first at night.
00:03:01.199 --> 00:03:06.639
Those gaps are bigger than the overall one and they point in opposite directions.
00:03:07.280 --> 00:03:10.479
Here is what caught me off guard when I dug into this.
00:03:10.719 --> 00:03:14.400
The biggest swing in this data is not between lines.
00:03:14.560 --> 00:03:15.759
It is one winter.
00:03:16.000 --> 00:03:23.199
In January 2026, morning peak trains on the Hudson and Harlem lines were on time only 84.4% of the time.
00:03:23.439 --> 00:03:25.280
That is about 16% late.
00:03:25.520 --> 00:03:27.360
New Haven was 87.8.
00:03:27.759 --> 00:03:33.520
In February, Hudson 89, Harlem 88.6, New Haven 87.5.
00:03:34.000 --> 00:03:37.039
All three lines were lower together, two months running.
00:03:37.199 --> 00:03:39.520
The data does not say why.
00:03:39.759 --> 00:03:44.400
Three lines moving at once points to something they share, but not what.
00:03:45.039 --> 00:03:48.240
We also checked the two winters before it and they look normal.
00:03:48.400 --> 00:03:52.719
If you rode in at rush hour last winter, you had a similar ride on any of the three.
00:03:52.960 --> 00:03:55.919
So last January all three lines had a rough morning.
00:03:56.159 --> 00:03:57.360
What this means for you.
00:03:57.599 --> 00:04:05.280
Over the past year, about ninety-seven of every hundred trains on any of these three lines finish within six minutes of schedule.
00:04:05.520 --> 00:04:16.000
If you board at Port Chester Rye or New Rochelle, you ride New Haven, Tarrytown or Croton Harmon, the Hudson, White Plains or Bedford Hills, the Harlem.
00:04:16.319 --> 00:04:20.480
Pick your town for the fares, the parking, and the walk to the station.
00:04:20.720 --> 00:04:37.439
The lines may still differ inside Westchester, and this data set cannot show it, so if your train is late, you are allowed to be annoyed, and keep one eye on this winter, because all three lines were lower together last January, and the data cannot say it will not return.
00:04:37.680 --> 00:04:44.079
What this cannot tell you, how crowded the train was, what you paid, or how late the late ones ran.
00:04:44.240 --> 00:04:47.600
It counts trains, not minutes, and not comfort.
00:04:48.160 --> 00:04:49.839
That is your Friday intel.
00:04:50.000 --> 00:04:53.360
If this was useful, share it with someone who lives here.
00:04:53.600 --> 00:04:56.319
I'll see you Monday on the Westchester Brief.
00:04:56.399 --> 00:04:58.240
I'm Jim, and I live here.