Southern California Mango Hobbyist Blog

60 Mangos, One Ranked List: A Data-Driven Guide for SoCal Growers

1,356 tastings. 60 varieties. One question: which mangos actually deserve a spot in your Southern California backyard?

๐Ÿฅญ Based on the Comprehensive Mango Tasting Analysis dataset  |  Ranked by Normalized Score

If you have ever agonized over which mango tree to plant next in your Tustin, Vista, or Encinitas backyard, this blog post is your cheat sheet. We took the full Comprehensive Mango Tasting Analysis dataset โ€” 60 varieties, more than 1,300 tastings โ€” and turned it into a ranked hobbyist guide, sorted primarily by the Normalized score that balances raw popularity with rating strength.

Why This Ranking Matters for SoCal Hobbyists

Southern California mango growers face an interesting problem: we have space for two, maybe three trees in a lifetime โ€” not fifty. Picking wrong means five to seven years before you realize you should have planted something else. This dataset gives you the collective palate of a large tasting community, condensed into four numbers per variety:

Dataset at a Glance

60
Varieties Ranked
1,356
Total Tastings
4.02
Top Normalized (Bombay)
2.90
Bottom Normalized (Bailey's Marvel)
78
Highest Sample Count (Duncan)
295
Highest Weighted Score (Sweet Tart)
"A high average with only four tastings is a rumor. A high normalized score with 60+ tastings is a recommendation."

The Elite Ten: Highest Normalized Scores

These are the ten varieties that sit at the very top of the ranking. If you are building a curated SoCal backyard collection and want to hedge toward crowd favorites, the shortlist starts here. Note that Bombay tops the chart on Normalized score even with only 10 tastings โ€” its 4.7 average is the strongest raw rating in the entire dataset.

RankVarietyAverageCountWeightedNormalized
1Bombay Top Score4.7010474.0174
2Fruit Punch4.504183.9177
3Coconut Cream Big Sample4.21662783.8470
4Sweet Tart Big Sample4.04732953.7742
5Orange Sherbet Big Sample3.91562193.6958
6Fruit Cocktail4.0013523.6933
7Peach Cobbler4.0010403.6902
8Karen Michelle3.87381473.6575
9Juicy Peach3.9213513.6573
10Dwarf Hawaiian3.9111433.6487

The Big Three Consensus Picks

When you filter for high Normalized scores AND large sample counts (40+ tastings), the dataset spotlights three names that are impossible to ignore. These varieties have been tasted by dozens of independent evaluators and still finished at or near the top โ€” the strongest signal in the entire dataset.

Coconut Cream #3 Overall

Average: 4.21 ยท Count: 66 ยท Weighted: 278 ยท Normalized: 3.85

The community darling. Two-thirds of the tastings scored it above 4, and its normalized score of 3.85 is the highest of any variety with 60+ tastings.

Sweet Tart #4 Overall

Average: 4.04 ยท Count: 73 ยท Weighted: 295 ยท Normalized: 3.77

The most-tasted top performer. Its 295 weighted score is the highest in the entire dataset โ€” meaning more total "vote value" than any other variety.

Orange Sherbet #5 Overall

Average: 3.91 ยท Count: 56 ยท Weighted: 219 ยท Normalized: 3.70

The third pillar of the well-sampled elite. It clears every high-confidence threshold and rounds out the trio.

Strong Performers (Normalized 3.50 โ€“ 3.68)

These varieties sit just under the elite tier but still cleared the 3.50 normalized threshold. That put twenty varieties total in what we would call the "highly-rated" band of the dataset. If space is tight, this is the pool most SoCal hobbyists should be shopping from.

RankVarietyAverageCountWeightedNormalized
11Pineapple Pleasure3.8614543.6275
12Sunrise3.7010373.5500
13Philippine3.6913483.5495
14Mallika3.58361293.5222
15Lemon Zest3.6015543.5084
15White Piri3.6015543.5084
17Pickering3.5917613.5049
18Carrie3.5812433.4975
19Zill3.569323.4814
20ST Maui3.5022773.4688

๐Ÿ’ก Hobbyist Tip: Watch for the Twins

Lemon Zest and White Piri sit at identical stats โ€” a 3.60 average across 15 tastings, both at 3.5084 normalized. It is worth noting when two varieties land on the exact same rating profile because it means the community sees them as effectively interchangeable in the "very solid" tier.

The Well-Sampled Workhorses

Sample size is a hobbyist's best friend. Below is the shortlist of every variety that received at least 40 tastings โ€” the ones that carry the most statistical weight regardless of where they landed on the ranking. Two varieties in this list finished the ranking below the median, which is itself an interesting data point.

RankVarietyCountAverageNormalizedSignal
22Duncan783.363.4605Largest sample, middling score
4Sweet Tart734.043.7742Highest weighted (295)
3Coconut Cream664.213.8470Elite + big sample
5Orange Sherbet563.913.6958Elite + big sample
24CAC543.393.4498Solid mid-tier
43Angie533.043.2846Well-sampled but lower half
27Lemon Meringue473.383.4398Solid mid-tier
35Simmonds473.193.3503Well-sampled but middling

The Mid-Tier Middle (Ranks 21 โ€“ 40)

The middle of the ranking is where the majority of well-known heritage varieties land. These are not bad mangos in the dataset โ€” they are simply competing against a very strong top ten. Notably, Alphonso, a variety with global reputation, lands at rank 40 on this dataset with a 3.13 average across 23 tastings.

RankVarietyAverageCountWeightedNormalized
21Edward3.46351213.4622
22Duncan3.36782623.4605
23Spirit of 763.5010353.4565
24CAC3.39541833.4498
25Sonpari3.4823803.4481
26Gouveia3.4613453.4416
27Lemon Meringue3.38471593.4398
28Gopher3.4224823.4319
29DOT3.3923783.4190
30P-223.3625843.4065
3136-433.3813443.4056
32Cat Hoa Loc3.3727913.3995
33Step3.3523773.3987
34Zill 33-83.3010333.3630
35Simmonds3.19471503.3503
36Yi Xuan3.2313423.3337
36Cogshell3.2313423.3337
38Rosigold3.1526823.3111
39Maha Chinook3.1513413.2978
40Alphonso3.1323723.2971

The Lower Half (Ranks 41 โ€“ 60)

These are the varieties the tasting community rated lowest. Most of them still cleared a 2.0 average, but they land in the bottom third of the ranking. If your goal is to maximize the eating quality of a limited number of trees, this is the tier to be most skeptical about.

RankVarietyAverageCountWeightedNormalized
41Jehangir3.1414443.2937
42Carla3.08391203.2886
43Angie3.04531613.2846
44Nam Doc Mai3.0824743.2761
45Naomi3.1010313.2695
46Brahm Kai Meu3.0714433.2603
47Cecilove3.0522673.2564
48Harumanis3.0013393.2258
48Juliette3.0013393.2258
50Pina Colada2.9315443.1967
51Maha Chanok2.9021613.1788
52Sugarloaf2.8614403.1601
53Gaylour2.8513373.1539
54Giselle2.8318513.1531
55Sturrock2.7823643.1345
56USDA S-102.754113.0997
57Maya2.7010273.0825
58PK2.6015393.0409
59Gold Nugget2.4314342.9598
60Bailey's Marvel2.3010232.8956

Patterns Worth Noticing

Pattern 1: Popularity Does Not Equal Praise

Duncan received 78 tastings โ€” more than any other variety โ€” yet finished at rank 22 with only a 3.36 average. High familiarity in the dataset did not translate into a top rating.

Pattern 2: Small-Sample Winners

Bombay (10 tastings) and Fruit Punch (4 tastings) hold the top two normalized scores. Their averages of 4.7 and 4.5 are the highest in the dataset, but sample sizes are small enough to warrant caution.

Pattern 3: The 3.5 Cliff

Twenty of the 60 varieties clear a 3.50 normalized score. Below rank 20, the ranking descends steadily into the mid-3.0s, indicating a clear performance cliff between "top tier" and "everything else."

Pattern 4: The Weighted-Score Champion

Sweet Tart posts the highest weighted score in the dataset (295), edging out Coconut Cream (278) and Duncan (262). It is the community's most heavily-voted top performer.

How SoCal Hobbyists Might Read This Data

The dataset is a snapshot of collective taste preference โ€” it is not a growing guide, disease-resistance chart, or climate suitability index. With that caveat clearly on the record, here is how a hobbyist can reasonably use these numbers:

If you want the safest crowd-pleaser

Lean toward the top-tier varieties with both a strong normalized score AND a large sample count: Coconut Cream, Sweet Tart, and Orange Sherbet. These are the only three varieties in the dataset with 50+ tastings and a normalized score above 3.69.

If you like a small-sample "sleeper" bet

Names like Bombay (4.70 avg), Fruit Cocktail (4.00 avg), Peach Cobbler (4.00 avg), Juicy Peach (3.92 avg), and Dwarf Hawaiian (3.91 avg) all posted averages of 3.9 or higher, but with fewer than 15 tastings each. Excellent shots on goal, but the dataset itself flags that there is less statistical certainty.

If you want a well-known variety with a track record

Look at Karen Michelle (rank 8, 38 tastings), Mallika (rank 14, 36 tastings), Edward (rank 21, 35 tastings), and Pickering (rank 17, 17 tastings). All are widely-tasted and land in the strong-to-mid tier.

If you are limiting trees and want to avoid disappointment

The bottom ten โ€” Maha Chanok, Sugarloaf, Gaylour, Giselle, Sturrock, USDA S-10, Maya, PK, Gold Nugget, and Bailey's Marvel โ€” all scored below 3.20 normalized. That is where the dataset registered the weakest community enthusiasm.

Closing Thoughts for the SoCal Hobbyist

The Southern California mango scene lives and dies by variety selection. With five to seven years between planting and reliable fruit, every mango hobbyist eventually asks the same question: am I growing what I actually want to eat? This dataset โ€” 60 varieties and more than 1,300 tastings condensed into four columns โ€” is one of the cleanest signals we have to answer that question.

The takeaways are simple. Coconut Cream, Sweet Tart, and Orange Sherbet are the closest thing to a data-backed consensus. Bombay is the highest-rated variety in the dataset by a comfortable margin, even with a modest sample count. And a surprisingly large group of well-known names โ€” Alphonso, Nam Doc Mai, Angie, Pina Colada โ€” land in the lower half of this particular ranking, which is a useful reality check against reputation alone.

Whatever you plant next, plant it deliberately. The trees you put in the ground today are the mangos your family eats a decade from now.

"60 varieties, 1,356 tastings, one honest ranking. Choose accordingly."

All data in this post is sourced exclusively from the Comprehensive Mango Tasting Analysis dataset. No external sources, growing references, or third-party rankings were used. Rankings are ordered by Normalized score; ties are noted where applicable.